A q u a nt it at iv e st udy of w or k - lif e exp er iences

in t he p ub lic a nd pr iv ate s ect ors of t he

A us tr a lia n const ruc t ion ind us tr y

Valerie Francis and Helen Lingard

Published by:
Construction Industry Institute, Australia Inc,
Queensland University of Technology
2 George Street
Brisbane, Queensland, 4000.
September 2004
 The University of Melbourne 2004
Executive Summary
The study was undertaken to provide a greater understanding of the work-life balance experiences of
project based managerial employees within the private and public sectors of the Australian construction
industry. In particular, the relationships between job and family characteristics, work-life conflict and
employees well-being, family functioning and work-related attitudes were investigated.

The work-life experiences of employees in the public and private sectors of the Australian construction
industry were quantified by means of a web-based survey. The results revealed that there were notable
differences between the responses of public and private sector employees on a number of variables of
interest. These are summarised below:

Private sector employees work significantly longer hours than public sector employees.

Public sector employees express a significantly lower turnover intent than private sector
employees.

Private sector employees have significantly less work-related flexibility than public sector
employees.

Private sector employees experience significantly higher levels of work interference with family
life than public sector employees.

Public sector employees find engaging in leisure activities with family and friends and taking
part in domestic and child rearing activities significantly less difficult than their public sector
counterparts.

Public sector employees demonstrate higher levels of continuance commitment than private
sector employees.
In comparison with international norm scores for the burnout dimensions, the levels of burnout
experienced by respondents in our sample were very high. The extent to which work-to-family conflict
mediates the relationship between job demands (work hours, responsibility and subjective workload)
and the emotional exhaustion dimension of burnout were investigated. The results from the current
study are consistent with those of other researchers who report that job demands affect emotional
exhaustion indirectly via work-to-family conflict. The implication of these findings is that the reduction
of work-to-family conflict, for example through the implementation of family-responsive management
interventions, may be an effective control measure for the risk of employee burnout.
Our results also clearly demonstrate the importance of a supportive work environment. Perceived
organisational support (POS), emotional support from ones supervisor, and emotional support
from ones co-workers all demonstrated significant main effects on the emotional exhaustion
dimension of employee burnout. In addition to these main effects, our results also suggest that POS
and supervisors and co-workers practical support for employees mitigate the ill-effects of work-to-
family conflict. The ability of a supportive work environment to buffer the effect of work-to-family
conflict on burnout suggests that organisational interventions designed to engender an organisational
culture that is supportive of employees work-life balance would be of benefit.
The study examined the role of flexibility or control that employees have over their work and tested the
extent to which this perceived control has a main effect on employees levels of work-to-family conflict
and/or moderates the relationship between work hours and work-to-family conflict. We found that
perceived control has a main effect on work-to-family conflict but does not moderate the relationship
between work hours and work-to-family conflict. We recommend that both organisations, but
particularly the private sector organisation, find ways to provide employees with a degree of control
over their work arrangements.
The integrative model of the work-family interface that was developed indicated that the non-work
domain is more susceptible to interference from work than the work domain is to interference from
family. Thus work-to-family conflict had a negative impact on family relationships, in particular, the
relationship with ones spouse/partner. However, work attitudes and/or distress were unaffected by

i
family-to-work conflict. This asymmetrical pattern of relationships suggests that employees in our
sample were better at segregating and protecting the work domain from family interference than vice-
versa. It also suggests that the employees in our sample were more likely to make adjustments in their
non-work lives to accommodate work demands.
The experiences of partners/spouses of a subset of employees in our sample (n=46) were examined.
Bi-variate correlations revealed that respondents who believed their relationship to be affected by their
partners (i.e. the employees) work commitments, reported significantly lower levels of companionship
with their partner, significantly higher relationship tension and significantly higher social undermining
behaviour by their partner. Respondents who consider their relationship is affected by their partners
(i.e. the employees) work commitments also report undertaking significantly more hours of household
work each week and significantly lower satisfaction with the sharing of household chores. Although the
sample size was too small to examine crossover effects between partners, these correlations indicate
that, in some households, employees work commitments have a detrimental impact upon family
functioning.
Preferences for work-life benefits were found to vary considerably between different groups of
employees in our sample. Childcare assistance support was highest among employees between 30-49
years of age and was significantly greater among employees who were partnered with dependent
children than those in other family structures. By comparison, wellness and personal development
benefits were highest among the 20-29 year age group and progressively declined with age. Wellness
and personal development benefits were most strongly preferred by single persons and partnered
employees without dependent children. The results suggest that no one work-life balance solution is
applicable to all employees. We recommend workforce profiling, by business unit or organisational sub-
unit might be a useful means to identify the most beneficial and highly valued initiatives to implement.
Recommendations arising from this stage of the project include:
1. Both organisations could seek reduce employee burnout by trying to address known antecedents of
burnout including:

Work hours

Subjective workload

Role conflict

Work-family conflict.
2. Both organisations should also focus on creating a work environment that is supportive of
employees work-life balance. In particular, strategies could be developed to:

Create a supportive organisational culture

Coach supervisors in supporting employee work-life balance

Foster work groups that are supportive of employee work-life balance.
3. The public sector organisation should address the issue of high continuance commitment among its
employees. The following could be considered as strategies:

Job enrichment or rotation, to ensure employees perceive that their skills and experience
remain current

Strategies to support managerial autonomy and decision-making

Professional development opportunities to ensure employees technical and managerial skills
remain up-to-date and in line with practices in the private sector.
4. The private sector organisation could address the following issues in particular:

Focus on the reduction of employees work-to-family conflict as a means of reducing turnover
intentions,

Develop strategies to improve employees control over their work arrangements;

Consider ways to reduce employees work hours, and

Develop a work culture more conducive to work-life balance.
The private sector could learn important lessons from the public sector organisation in all of these
areas.

ii
Acknowledgements

The authors would sincerely like to thank the Department of Main Roads Queensland for funding this
study and for the Construction Industry Institute, Australia for administering the grant.

In particular we would also like to thank members of the project steering committee from DMR (Qld),
Baulderstone Hornibrook, Leighton Contractors, Theiss and Transfield, as well as the employees, and
their spouse/partners from the public and private sector organisations surveyed as part of this study.

Lastly we would like to thank Rochelle Cairns and Karmen Jobling for their research assistance.

iii
Table of Contents

EXECUTIVE SUMMARY ...............................................................................................................................I

ACKNOWLEDGEMENTS ................................................................................................................................... III
TABLE OF CONTENTS ..................................................................................................................................... IV
LIST OF TABLES ........................................................................................................................................... VIII
LIST OF FIGURES ............................................................................................................................................. X
SECTION 1: INTRODUCTION ..................................................................................................................... 1
REPORT OVERVIEW ......................................................................................................................................... 1
BACKGROUND ................................................................................................................................................. 1
DEFINING WORK-LIFE BALANCE ...................................................................................................................... 3
Work-life or work-family?.......................................................................................................................... 3
THE CONSTRUCTION INDUSTRY ....................................................................................................................... 3
THE STUDY OBJECTIVES .................................................................................................................................. 4
CONCEPTUAL FRAMEWORK ............................................................................................................................. 4
Work-life conflict........................................................................................................................................ 4
ANTECEDENTS OF WORK-LIFE CONFLICT ......................................................................................................... 6
ORGANISATIONAL ANTECEDENTS OF WORK-LIFE CONFLICT ............................................................................ 6
Job demands............................................................................................................................................... 6
FAMILY ANTECEDENTS OF WORK-LIFE CONFLICT ............................................................................................ 8
Family responsibilities ............................................................................................................................... 8
OUTCOMES OF WORK-LIFE CONFLICT .............................................................................................................. 9
Individual outcomes ................................................................................................................................... 9
Family outcome variables ........................................................................................................................ 10
Organisational outcome variables ........................................................................................................... 11
MODERATOR VARIABLES .............................................................................................................................. 13
Organisational and social support........................................................................................................... 13
Flexible work arrangements .................................................................................................................... 15
WORK-LIFE BENEFITS .................................................................................................................................... 16
PUBLIC/PRIVATE SECTOR DIFFERENCES ......................................................................................................... 17
DOMESTIC PARTNERS .................................................................................................................................... 18
The effect of partners career/job involvement ........................................................................................ 18
Crossover effects.................................................................................................................................. 19
SECTION 2: METHODOLOGY .................................................................................................................. 21
DATA COLLECTION AND SAMPLING ............................................................................................................... 21
Internet survey.......................................................................................................................................... 21
QUESTIONNAIRE DESIGN ............................................................................................................................... 22
EMPLOYEE QUESTIONNAIRE ......................................................................................................................... 23
Part A Demographic information ...................................................................................................... 23
Part B Work load and responsibility.................................................................................................. 23
Part C Your work environment .......................................................................................................... 24
Part D Your feelings about work........................................................................................................ 25
Part E Spouse/partner relationship.................................................................................................... 26
Part F Family dependants.................................................................................................................. 27
Part G Absence from work ................................................................................................................. 28
Part H Work-life balance ................................................................................................................... 28
SPOUSE/PARTNERS QUESTIONNAIRE ............................................................................................................ 30
Part A Demographic information ...................................................................................................... 30
Part B Your employment .................................................................................................................... 30
Part C Work-life balance ................................................................................................................... 31
Part D Spouse/partner relationship.................................................................................................... 31
Part E Family dependants.................................................................................................................. 31
Part F Household management ......................................................................................................... 31
DATA ANALYSIS ............................................................................................................................................ 32
Reliability of measurement....................................................................................................................... 32
 Factor analysis......................................................................................................................................... 32
Statistical comparisons of means ............................................................................................................. 33
Correlation analysis................................................................................................................................. 33
Multiple regression analysis .................................................................................................................... 33
Moderators and mediators....................................................................................................................... 34
SECTION 3: RESULTS ................................................................................................................................. 37
INTRODUCTION.............................................................................................................................................. 37
EMPLOYEE QUESTIONNAIRE .......................................................................................................................... 37
Demographic information........................................................................................................................ 37
Work related information......................................................................................................................... 40
Family-related information...................................................................................................................... 50
Absence from work................................................................................................................................... 53
Work load and responsibility for things and persons............................................................................... 55
Support at work........................................................................................................................................ 56
Feelings about work................................................................................................................................. 58
Relationship with spouse/partner............................................................................................................. 63
Parenting issues ....................................................................................................................................... 64
Work-life balance..................................................................................................................................... 66
Work-life benefits ..................................................................................................................................... 69
COMPARISON BETWEEN THE PUBLIC AND PRIVATE SECTORS ......................................................................... 73
Work load and responsibility ................................................................................................................... 73
Support at work........................................................................................................................................ 74
Feelings about work................................................................................................................................. 77
Burnout .................................................................................................................................................... 78
Spouse/partner relationship ..................................................................................................................... 79
Work-life balance..................................................................................................................................... 79
Summary of public and private sector employee differences ................................................................... 82
BI-VARIATE CORRELATIONS FOR EMPLOYEE DATA........................................................................................ 83
Whole sample ........................................................................................................................................... 83
Summary of bi-variate correlation results ............................................................................................... 85
Partnered employees................................................................................................................................ 86
Employees who are parents of dependent children.................................................................................. 88
BURNOUT ANALYSIS ..................................................................................................................................... 90
Work-family conflict as a mediator of the job stress-burnout relationship.............................................. 91
JOB CONTROL/FLEXIBILITY............................................................................................................................ 94
SUPPORTIVE WORK ENVIRONMENTS, WORK-TO-FAMILY CONFLICT AND BURNOUT ....................................... 96
Perceived organisational support ............................................................................................................ 96
Supervisor support ................................................................................................................................... 97
Co-worker support ................................................................................................................................. 101
THE NATURE, ANTECEDENTS AND OUTCOMES OF WORK-FAMILY CONFLICT ................................................ 103
Cross-role effects ................................................................................................................................... 103
Within-role effects .................................................................................................................................. 106
RESULTS OF PARTNERS QUESTIONNAIRE .................................................................................................... 108
Introduction............................................................................................................................................ 108
Demographic Information...................................................................................................................... 108
Employment............................................................................................................................................ 110
Work-life control .................................................................................................................................... 114
Relationship with spouse/partner........................................................................................................... 114
Family dependants ................................................................................................................................. 116
Household management......................................................................................................................... 118
BI-VARIATE CORRELATIONS FOR SPOUSE/PARTNER ..................................................................................... 122
Whole sample ......................................................................................................................................... 122
Partners who work ................................................................................................................................. 122
SECTION 4: DISCUSSION AND CONCLUSIONS ................................................................................. 125
PUBLIC/PRIVATE SECTOR DIFFERENCES ....................................................................................................... 125
Private sector ......................................................................................................................................... 125
Public sector .......................................................................................................................................... 125
 BURNOUT .................................................................................................................................................... 126
SUPPORTIVE WORK ENVIRONMENT.............................................................................................................. 128
FLEXIBILITY AND CONTROL ........................................................................................................................ 129
WORK-TO-FAMILY AND FAMILY-TO-WORK CONFLICT ................................................................................. 130
WORK-FAMILY PREFERENCES ..................................................................................................................... 131
RECOMMENDATIONS ................................................................................................................................... 131
Public and Private sector organisations................................................................................................ 131
Public sector .......................................................................................................................................... 132
Private sector ......................................................................................................................................... 132
REFERENCES ............................................................................................................................................. 135
List of Tables

TABLE 1 RESPONSIBILITY FOR DEPENDANTS (RFD) MEASURE ITEM AND ITEM WEIGHTS ........................ 27
TABLE 2 DEMOGRAPHIC CHARACTERISTICS OF SAMPLE............................................................................. 38
TABLE 3 POSITION DESCRIPTION BY SAMPLE, COMPANY AND GENDER ...................................................... 44
TABLE 4 WORK LOCATION ......................................................................................................................... 45
TABLE 5 PROJECT CONSTRUCTION COST BY COMPANY AND GENDER ......................................................... 47
TABLE 6 FAMILY STRUCTURE BY COMPANY AND GENDER ......................................................................... 50
TABLE 7 SPOUSE/PARTNER STATUS ............................................................................................................ 51
TABLE 8 NUMBER OF CHILDREN. ............................................................................................................... 52
TABLE 9 SUBJECTIVE QUANTITATIVE WORK LOAD MEAN ITEM SCORES ..................................................... 55
TABLE 10 RESPONSIBILITY FOR THINGS MEAN ITEM SCORES ....................................................................... 55
TABLE 11 RESPONSIBILITY FOR PEOPLE MEAN ITEM SCORES ....................................................................... 56
TABLE 12 ORGANISATIONAL SUPPORT MEAN ITEM SCORES ......................................................................... 56
TABLE 13 FACTOR ANALYSIS FOR SUPERVISOR SUPPORT ............................................................................. 57
TABLE 14 SUPERVISOR SUPPORT MEAN SCORES ........................................................................................... 57
TABLE 15 CO-WORKER SUPPORT (PRACTICAL) MEAN ITEM SCORES ............................................................. 58
TABLE 16 CO-WORKER SUPPORT (EMOTIONAL) MEAN ITEM SCORES ............................................................ 58
TABLE 17 FACTOR ANALYSIS FOR ORGANISATIONAL COMMITMENT ............................................................ 59
TABLE 18 ORGANISATIONAL COMMITMENT MEAN SCORES .......................................................................... 59
TABLE 19 TURNOVER INTENT MEAN ITEM SCORES ....................................................................................... 60
TABLE 20 JOB SATISFACTION MEAN ITEM SCORES........................................................................................ 60
TABLE 21 JOB SATISFACTION MEAN SCORES ................................................................................................ 60
TABLE 22 BURNOUT MEAN ITEM SCORES ..................................................................................................... 61
TABLE 23 FACTOR ANALYSIS FOR BURNOUT ................................................................................................ 62
TABLE 24 BURNOUT MEAN SCORES.............................................................................................................. 62
TABLE 25 RELATIONSHIP SOCIABILITY MEAN ITEM SCORES ......................................................................... 63
TABLE 26 COMPANIONSHIP WITH PARTNER MEAN ITEM SCORES .................................................................. 63
TABLE 27 RELATIONSHIP TENSION MEAN ITEM SCORES ............................................................................... 64
TABLE 28 SOCIAL UNDERMINING MEAN ITEM SCORES ................................................................................. 64
TABLE 29 PARENTING BEHAVIOURS MEAN ITEM SCORES ............................................................................. 65
TABLE 30 PARENTAL ATTITUDES MEAN SCORES .......................................................................................... 65
TABLE 31 SUPPORT WITH DEPENDANTS ....................................................................................................... 66
TABLE 32 WORK-TO-FAMILY CONFLICT MEAN ITEM SCORES ....................................................................... 66
TABLE 33 FAMILY-TO-WORK CONFLICT MEAN ITEM SCORES ....................................................................... 66
TABLE 34 FACTOR ANALYSIS FOR FAMILY MANAGEMENT ........................................................................... 67
TABLE 35 FAMILY MANAGEMENT MEAN SCORES ......................................................................................... 68
TABLE 36 FACTOR ANALYSIS FOR WORK-LIFE CONTROL.............................................................................. 68
TABLE 37 WORK-LIFE CONTROL MEAN SCORES ........................................................................................... 69
TABLE 38 WORK-LIFE BENEFITS MEAN ITEM SCORES ................................................................................... 69
TABLE 39 FACTOR ANALYSIS FOR WORK-LIFE BENEFITS .............................................................................. 70
TABLE 40 WORK-LIFE BENEFIT PREFERENCES MEAN SCORES....................................................................... 70
TABLE 41 WORK-LIFE BENEFIT PREFERENCES MEAN SCORE FOR VARIOUS AGE COHORTS............................ 71
TABLE 42 WORK-LIFE BENEFIT PREFERENCES MEAN SCORE FOR DIFFERENT FAMILY STRUCTURES ............. 71
TABLE 43 BI-VARIATE CORRELATIONS BETWEEN WORK VARIABLES MEASURED IN THE STUDY ................... 84
TABLE 44 BI-VARIATE CORRELATIONS FOR PARTNERED EMPLOYEES .......................................................... 87
TABLE 45 BI-VARIATE CORRELATIONS FOR PARENTS OF DEPENDENT CHILDREN ......................................... 89
TABLE 46 CROSS OCCUPATIONAL COMPARISON OF MEAN BURNOUT SCORES ............................................... 90
TABLE 47 REGRESSION ANALYSIS EXAMINING WORK-TO-FAMILY CONFLICT AS A MEDIATOR IN THE WORK
HOURS-EMOTIONAL EXHAUSTION RELATIONSHIP ........................................................................ 92
TABLE 48 REGRESSION ANALYSIS EXAMINING WORK-TO-FAMILY CONFLICT AS A MEDIATOR IN THE
SUBJECTIVE QUANTITATIVE WORKLOADEMOTIONAL EXHAUSTION RELATIONSHIP.................... 92
TABLE 49 REGRESSION ANALYSIS EXAMINING WORK-TO-FAMILY CONFLICT AS A MEDIATOR IN THE
RESPONSIBILITY (FOR THINGS)-EMOTIONAL EXHAUSTION RELATIONSHIP ................................... 93
TABLE 50 REGRESSION ANALYSIS EXAMINING WORK-TO-FAMILY CONFLICT AS A MEDIATOR IN THE
RESPONSIBILITY (FOR PERSONS)-EMOTIONAL EXHAUSTION RELATIONSHIP ................................. 93
TABLE 51 MULTIPLE REGRESSION FOR WORK HOURS AND WORK-LIFE CONTROL AS PREDICTORS OF WORK-
TO-FAMILY CONFLICT .................................................................................................................. 95

viii
TABLE 52 MULTIPLE REGRESSION FOR WORK HOURS AND WORK-FAMILY CONTROL AS PREDICTORS OF
WORK-TO-FAMILY CONFLICT ....................................................................................................... 95
TABLE 53 MULTIPLE REGRESSION FOR WORK-TO-FAMILY CONFLICT AND POS AS PREDICTORS OF
EMOTIONAL EXHAUSTION (BURNOUT)......................................................................................... 97
TABLE 54 MULTIPLE REGRESSION FOR WORK-TO-FAMILY CONFLICT AND SUPERVISOR EMOTIONAL SUPPORT
AS PREDICTORS OF EMOTIONAL EXHAUSTION .............................................................................. 98
TABLE 55 MULTIPLE REGRESSION FOR WORK-TO-FAMILY CONFLICT AND SUPERVISOR PRACTICAL SUPPORT
AS PREDICTORS OF EMOTIONAL EXHAUSTION .............................................................................. 99
TABLE 56 MULTIPLE REGRESSION FOR WORK-TO-FAMILY CONFLICT AND SUPERVISOR
CRITICISM/RESENTMENT AS PREDICTORS OF EMOTIONAL EXHAUSTION ..................................... 100
TABLE 57 MULTIPLE REGRESSION FOR WORK-TO-FAMILY CONFLICT AND CO-WORKER EMOTIONAL SUPPORT
AS PREDICTORS OF EMOTIONAL EXHAUSTION ............................................................................ 101
TABLE 58 MULTIPLE REGRESSION FOR WORK-TO-FAMILY CONFLICT AND CO-WORKER PRACTICAL SUPPORT
AS PREDICTORS OF EMOTIONAL EXHAUSTION ............................................................................ 102
TABLE 59 DEMOGRAPHIC CHARACTERISTICS OF SAMPLE........................................................................... 108
TABLE 60 WORK ROLE, GENDER AND PARTNERS EMPLOYMENT SECTOR .................................................. 111
TABLE 61 WORK INVOLVEMENT ................................................................................................................ 113
TABLE 62 BURNOUT .................................................................................................................................. 113
TABLE 63 MEAN SCORE FOR BURNOUT ...................................................................................................... 114
TABLE 64 WORK-LIFE CONTROL ................................................................................................................ 114
TABLE 65 RELATIONSHIP SOCIABILITY ...................................................................................................... 115
TABLE 66 COMPANIONSHIP WITH PARTNER ............................................................................................... 115
TABLE 67 RELATIONSHIP TENSION ............................................................................................................. 115
TABLE 68 SOCIAL UNDERMINING ............................................................................................................... 116
TABLE 69 PARENTAL ATTITUDES OF RESPONDENTS ................................................................................... 117
TABLE 70 RESPONDENTS VIEWS OF PARENTAL ATTITUDES OF THEIR PARTNERS ....................................... 117
TABLE 71 MEAN SCORE FOR PARENTING ATTITUDES ................................................................................. 118
TABLE 72 MEAN SCORE FOR HOURS/DAY ON HOUSEHOLD ACTIVITIES ....................................................... 118
TABLE 73 WHO UNDERTAKES HOUSEHOLD ACTIVITIES .............................................................................. 119
TABLE 74 MEAN SCORE FOR FREQUENCY OF HOUSEHOLD ACTIVITIES ....................................................... 119
TABLE 75 MEAN SCORE FOR HOURS PERFORMED IN SPECIFIED TASK IN AN AVERAGE WEEK ...................... 120
TABLE 76 SATISFACTION WITH SHARING OF HOUSEHOLD CHORES ............................................................. 121
TABLE 77 SCHEDULE COMPATIBILITY ........................................................................................................ 121
TABLE 78 BI-VARIATE CORRELATIONS BETWEEN FAMILY AND PARTNERS WORK VARIABLES FOR WHOLE
SAMPLE ..................................................................................................................................... 122
TABLE 79 BI-VARIATE CORRELATIONS BETWEEN FAMILY AND PARTNERS WORK VARIABLES FOR PARENTS
WITHIN THE SAMPLE .................................................................................................................. 123
List of Figures

FIGURE 1 MEDIATING VARIABLE Z AND W (SOURCE, COHEN ET AL., 2003, P. 458). ................................... 34
FIGURE 2 CONTINGENT EFFECTS, Z AS A MODERATOR (SOURCE, COHEN ET AL., 2003, P. 458)................... 35
FIGURE 3 AGE COHORTS OF ALL RESPONDENTS ........................................................................................... 38
FIGURE 4 PUBLIC AND PRIVATE SECTOR EMPLOYEES BY AGE COHORTS ...................................................... 39
FIGURE 5 MALE AND FEMALE RESPONDENTS BY COMPANY ........................................................................ 39
FIGURE 6 MALE AND FEMALE RESPONDENTS BY AGE COHORTS .................................................................. 40
FIGURE 7 AUSTRALIAN-BORN RESPONDENTS BY COMPANY ........................................................................ 40
FIGURE 8 CONSTRUCTION EXPERIENCE BY COMPANY ................................................................................. 41
FIGURE 9 CONSTRUCTION EXPERIENCE BY GENDER .................................................................................... 41
FIGURE 10 CONSTRUCTION EXPERIENCE AND TENURE BY COMPANY ............................................................ 42
FIGURE 11 BREAKDOWN OF AVERAGE HOURS WORKED PER WEEK................................................................ 42
FIGURE 12 AVERAGE HOURS WORKED PER WEEK BY COMPANY .................................................................... 43
FIGURE 13 AVERAGE HOURS WORKED PER WEEK BY GENDER ....................................................................... 43
FIGURE 14 POSITION DESCRIPTION BY COMPANY .......................................................................................... 44
FIGURE 15 POSITION DESCRIPTION BY GENDER ............................................................................................. 45
FIGURE 16 WORK LOCATION BY COMPANY ................................................................................................... 46
FIGURE 17 WORK LOCATION BY GENDER ...................................................................................................... 46
FIGURE 18 TYPE OF PROJECT BY COMPANY. .................................................................................................. 47
FIGURE 19 PROJECT COMPLEXITY ................................................................................................................. 48
FIGURE 20 PROJECT TECHNICAL COMPLEXITY BY COMPANY......................................................................... 48
FIGURE 21 PROJECT TECHNICAL COMPLEXITY BY GENDER............................................................................ 49
FIGURE 22 FUTURE PROJECT PREFERENCE BY COMPANY ............................................................................... 49
FIGURE 23 FUTURE PROJECT PREFERENCE BY GENDER .................................................................................. 50
FIGURE 24 FAMILY STRUCTURE BY GENDER.................................................................................................. 50
FIGURE 25 SPOUSE/PARTNER IN PAID WORK BY COMPANY ............................................................................ 51
FIGURE 26 SPOUSE/PARTNER IN PAID WORK BY GENDER ............................................................................... 51
FIGURE 27 HOURS WORKED BY SPOUSE/PARTNER PER WEEK ........................................................................ 52
FIGURE 28 HOURS WORKED BY SPOUSE/PARTNER PER WEEK BY GENDER ..................................................... 52
FIGURE 29 NUMBER OF CHILDREN BY COMPANY ........................................................................................... 53
FIGURE 30 NUMBER OF CHILDREN BY GENDER.............................................................................................. 53
FIGURE 31 DAYS OFF IN PAST 12 MONTHS BY COMPANY ............................................................................... 54
FIGURE 32 ANNUAL LEAVE TAKEN BY COMPANY .......................................................................................... 54
FIGURE 33 WORK-LIFE BENEFIT PREFERENCES FOR VARIOUS AGE COHORTS ................................................. 71
FIGURE 34 WORK-LIFE BENEFIT PREFERENCES MEAN SCORE FOR DIFFERENT FAMILY STRUCTURES ............. 72
FIGURE 35 WORK-LIFE BENEFIT PREFERENCES MEAN SCORE BY GENDER ..................................................... 72
FIGURE 36 COMPARISON OF MEAN SCORES OF SUBJECTIVE QUALITATIVE WORKLOAD FOR PUBLIC AND
PRIVATE SECTOR ......................................................................................................................... 73
FIGURE 37 COMPARISON OF MEAN SCORES FOR RESPONSIBILITY FOR THINGS AND PERSONS FOR THE PUBLIC
AND PRIVATE SECTOR .................................................................................................................. 74
FIGURE 38 COMPARISON OF MEAN WEEKLY WORK HOURS FOR PUBLIC AND PRIVATE SECTOR ...................... 74
FIGURE 39 COMPARISON OF MEAN PERCEIVED ORGANISATIONAL SUPPORT FOR PUBLIC AND PRIVATE SECTOR
.................................................................................................................................................... 75
FIGURE 40 COMPARISON OF THE MEANS OF THE THREE DIMENSIONS OF SUPERVISORY SUPPORT FOR THE
PUBLIC AND PRIVATE SECTOR...................................................................................................... 75
FIGURE 41 COMPARISON OF MEANS FOR PRACTICAL CO-WORKER SUPPORT FOR PUBLIC AND PRIVATE SECTOR
RESPONDENTS ............................................................................................................................. 76
FIGURE 42 PUBLIC AND PRIVATE SECTOR MEANS FOR AFFECTIVE CO-WORKER SUPPORT .............................. 76
FIGURE 43 COMPARISON OF MEAN ORGANISATIONAL COMMITMENT SCORES FOR PUBLIC AND PRIVATE
SECTOR RESPONDENTS ................................................................................................................ 77
FIGURE 44 COMPARISON OF PUBLIC AND PRIVATE SECTOR MEAN TURNOVER INTENTION SCORES ................ 77
FIGURE 45 COMPARISON OF PUBLIC AND PRIVATE SECTOR MEAN SCORES FOR JOB SATISFACTION ............... 78
FIGURE 46 PUBLIC AND PRIVATE SECTOR MEAN BURNOUT SCORES............................................................... 78
FIGURE 47 PUBLIC AND PRIVATE SECTOR MEAN SCORES FOR COMPATIBILITY WITH PARTNER, RELATIONSHIP
SOCIABILITY AND RELATIONSHIP TENSION .................................................................................. 79
FIGURE 48 PUBLIC AND PRIVATE SECTOR MEAN SCORES FOR WORK-TO-FAMILY AND FAMILY-TO-WORK
CONFLICT SCORES ....................................................................................................................... 80

x
FIGURE 49 PUBLIC AND PRIVATE SECTOR MEAN SCORES FOR FAMILY MANAGEMENT SCORES ...................... 80
FIGURE 50 PUBLIC AND PRIVATE SECTOR MEAN SCORES FOR THE TWO DIMENSIONS OF CONTROL/FLEXIBILITY
(WORK AND CHILDREN)............................................................................................................... 81
FIGURE 51 PUBLIC AND PRIVATE SECTOR MEAN SCORES FOR THE FOUR DIMENSIONS OF WORK-LIFE BENEFIT.
.................................................................................................................................................... 81
FIGURE 52 MEAN SCORES FOR THE THREE DIMENSIONS OF BURNOUT ACCORDING TO JOB CLASSIFICATION . 90
FIGURE 53 WORK-TO-FAMILY CONFLICT AS A MEDIATOR OF THE JOB STRESS-BURNOUT RELATIONSHIP ...... 91
FIGURE 54 JOB CONTROL/FLEXIBILITY AS A MODERATOR OF THE RELATIONSHIP BETWEEN JOB DEMANDS AND
WORK-TO-FAMILY CONFLICT ....................................................................................................... 94
FIGURE 55 POS AS A MODERATOR OF THE RELATIONSHIP BETWEEN WORK-TO-FAMILY CONFLICT AND
EMOTIONAL EXHAUSTION............................................................................................................ 96
FIGURE 56 THE MODERATING EFFECT OF PERCEIVED ORGANIZATIONAL SUPPORT ON THE RELATIONSHIP
BETWEEN WORK TO FAMILY CONFLICT AND EMOTIONAL EXHAUSTION (BURNOUT) .................... 97
FIGURE 57 SUPERVISOR SUPPORT AS A MODERATOR OF THE RELATIONSHIP BETWEEN WORK TO FAMILY
CONFLICT AND EMOTIONAL EXHAUSTION.................................................................................... 98
FIGURE 58 THE MODERATING EFFECT OF PRACTICAL SUPPORT FROM ONES SUPERVISOR ON THE
RELATIONSHIP BETWEEN WORK TO FAMILY CONFLICT AND EMOTIONAL EXHAUSTION (BURNOUT)
.................................................................................................................................................... 99
FIGURE 59 THE MODERATING EFFECT OF SUPERVISOR CRITICISM/RESENTMENT OF EMPLOYEES FAMILY
OBLIGATIONS ON THE RELATIONSHIP BETWEEN WORK TO FAMILY CONFLICT AND EMOTIONAL
EXHAUSTION (BURNOUT) .......................................................................................................... 100
FIGURE 60 CO-WORKER SUPPORT AS A MODERATOR OF THE RELATIONSHIP BETWEEN WORK-TO-FAMILY
CONFLICT AND EMOTIONAL EXHAUSTION.................................................................................. 101
FIGURE 61 THE MODERATING EFFECT OF CO-WORKER PRACTICAL SUPPORT ON THE RELATIONSHIP BETWEEN
WORK TO FAMILY CONFLICT AND EMOTIONAL EXHAUSTION (BURNOUT) .................................. 102
FIGURE 62 CONCEPTUAL MODEL OF THE WORK-FAMILY INTERFACE (ADAPTED FROM FRONE ET AL. 1997) 104
FIGURE 63 SUMMARY OF STANDARDISED PATH CO-EFFICIENTS FOR WORK-FAMILY CONFLICT MODEL ....... 105
FIGURE 64 ALTERNATIVE MODEL OF WITHIN-DOMAIN ANTECEDENTS AND OUTCOMES OF WORK-TO-FAMILY
CONFLICT SHOWING STANDARDISED PATH COEFFICIENTS ......................................................... 107
FIGURE 65 AGE COHORTS OF ALL RESPONDENTS ......................................................................................... 109
FIGURE 66 RESPONDENTS AGE BY PARTNERS SECTOR OF EMPLOYMENT................................................... 109
FIGURE 67 PERCENTAGE OF MALE AND FEMALE RESPONDENTS BY PARTNERS EMPLOYMENT SECTOR ....... 110
FIGURE 68 MALE AND FEMALE RESPONDENTS BY AGE COHORTS ................................................................ 110
FIGURE 69 HOURS WORKED IN PAID EMPLOYMENT PER WEEK..................................................................... 111
FIGURE 70 DAYS OFF FOR HOLIDAYS........................................................................................................... 112
FIGURE 71 HOURS SPENT ON WORK AND HOUSEHOLD TASKS BY PARENTAL STATUS................................... 120
FIGURE 72 VARIOUS WORK-LIFE FACTORS AND PARENTAL STATUS ............................................................ 121
Section 1: Introduction
Report Overview

This report is presented in four main sections.

Section 1 provides an introduction to the report and then presents the background and rationale for
the work.

Section 2 presents the methodology employed including scales to be used for the measurement of
work-life conflict as well as antecedent and outcome variables.

Section 3 contains the results of the quantitative data analyses of the employee and domestic partner
surveys. The section begins with a general overview of the employee data and then presents
comparisons by industry sector, bi-variate associations and some regression analyses. The section
concludes with an analysis of the spouse/partner data.

Section 4 is the final section and includes a discussion of the results, conclusions and
recommendations.

Background

Striking changes to the Australian workforce composition have occurred since the 1950s. The roles and
expectations of women and men have changed significantly over the past 50 years. One consequence
has been a shift in the importance of work relative to family and leisure. More women than ever before
are in the workforce, reflecting rising educational levels, changing societal attitudes and declining birth
rates. In Australia, the employment participation rate of women rose from 36 per cent in 1968 to 54 per
cent in 1998 (Australian Bureau of Statistics 1998 & 1994). Associated with this is the unprecedented
increase in number of dual earner households, which has increased the likelihood that both male and
female workers will have domestic responsibilities in addition to work responsibilities.

Managing work and family responsibilities can be very difficult for employees in dual income families
and there is evidence that women are choosing to delay or forego child-bearing to pursue fulfilling
careers. The average age of first time mothers in Australia is 29 (The Age, 31 May 2003) and, according
to federal Minister for Ageing, Kevin Andrews, Australias declining fertility rate has been below its
population replacement rate for 20 years. It is now at an all-time low of 1.73 births per woman of child-
bearing age (The Age, 31 May 2003). The trend to delay child bearing means that an increasing
proportion of couples will face the additional responsibility of dependent children and care of elderly
parents simultaneously or sequentially. The recent change from institutional aged care to home and
community-based care means that responsibility for caring for elderly relatives now rests with family
members. With Australias aging population and increasing life expectancy, the number of workers with
elder care responsibilities is likely to rise. In fact it is predicted that, between 1996 and 2041, the aged
dependency ratio will double from 18.1 to 34.8. This means that for every 100 workers there will be 34
aged dependants (Bourke, 2000).

Research highlights that issues of work-life balance are having serious impacts on social and educational
issues in Australia. A recent report by the Australian Childhood Foundation reports that almost two-
thirds of Australian parents surveyed believe they could be better parents. In particular, parents
identified the expectations and pace of workplaces impeded their ability to fulfill their parenting role.
Three-quarters of parents interviewed believed balancing work and family was a serious issue for them
and 71% of parents struggled to find the time to enjoy activities with their children (Tucci, Goddard &
Mitchell, 2004). The report concludes that parents lack of time for children is a major failing in

1
Australian society and that family-responsive workplaces must become a priority for public policy and
action by all levels of business and the community (Tucci et al., 2004). In another study undertaken by
the Australian Institute of Family Studies, children were asked to comment on their experience of their
parents work patterns. Most children, particularly of primary school age, indicated that they wanted
their parents to be more involved in aspects of their lives, for example, attending school events or
picking them up from school (Lewis, Hand & Tudball, 2001).

In his book, Crowded Lives, Lindsay Tanner describes the ability of workers to integrate the needs of
their family relationships with their work obligations as one of the central issues of twenty-first century
politics. He writes of the importance of enabling people to manage their relationships better and
emphasises that this not only applies to women but that men should be able to play a greater role in
parenting and family life (Tanner, 2003).

An Australian study by Russell, Barclay, Edgecombe, Donovan, Habib, Callaghan, and Pawson (1999)
found that 57% fathers, on being asked to report on the major barrier to them becoming involved as a
father, indicated work-related factors (30% noted work load and work commitments, 16% time
pressures and 11% having to earn an income). A large proportion of working fathers also experienced
high levels of stress. The report strongly supported the development of strategies specifically for men
to help them better balance their work and family responsibilities.

The consequences of work-life imbalance can be dire in terms of the profitability of organisations.
Organisational psychologists have shown that high levels of work-life balance stress can significantly
increase levels of absenteeism and turnover, and substantially decrease employees levels of job
satisfaction, job performance and commitment to the organisation. Hobson, Delunas and Kesic (2001)
documented links between a failure to achieve balance in employees work and personal lives with
increased absenteeism and lower levels of productivity and performance. Perhaps more significantly,
Allen (1999) determined that a reduction in levels of work-life balance strain can result in decreased
employee turnover, improved job satisfaction and increased levels of organisational commitment.

The failure to meet the work-life balance expectations of the workforce also poses serious economic
consequences, threatening Australias long-term economic growth. Indeed, Australias national labour
shortage is forecast to exceed 500,000 by 2020 (The Age, 26 June, 2003). Industries and organisations
that fail to meet employees work-life balance expectations face high rates of turnover and labour
shortages.

Both sides of Australian politics agree that work-life balance is an issue of critical importance to
Australia. Prime Minister John Howard has described the need to balance work and family as the most
important social issue facing Australia and, in 2003, the situation led the Australian Council of Trade
Unions (ACTU) to lodge a landmark test case with the Australian Industrial Relations Commission.
ACTU president Sharan Burrow noted that the test case was introduced to bring about a cultural
shift in the way Australians look at work (reported in The Australian, 25 June, 2003). The test case
sought to:

Extend unpaid parental leave for children under two years of age;

Allow eight weeks of simultaneous unpaid parental leave;

Allow part-time return to work after parental leave until a child is of school age; and

Give employees the right to request to change their work hours to accommodate family needs.

Despite the unanimous agreement that work-life balance is an issue of great social and economic
importance facing Australia in the 21st Century, there is less consensus about how the issue should best
be addressed. Nor has the magnitude of the problem been fully acknowledged in some industry sectors.

There is a need to gain a better understanding of the experiences and expectations of employees, as well
as their domestic partners, in an attempt to identify the source of the problem and suggest ways in
which organisations can assist their employees to achieve a better work-life balance.

2
Defining work-life balance

Work-life balance has received much attention in the popular management literature and media. While
much has been written about work-life balance, most studies do not explicitly define the concept. In
some instances work-life balance is used as a noun, for example when employees are encouraged to
find a balance. In other instances it is treated as a verb, as in to balance work and other aspects of
life.

Recently, Greenhaus, Collins and Shaw (2003) proposed a robust definition of work-family balance in
an attempt to distinguish it from other concepts. They defined work-family balance as the extent to
which an individual is equally engaged in and equally satisfied with his or her work role and family
role (Greenhaus et al., 2003, p.513). This definition does not view balance as universally good, but
instead recognises that balance can be either positive or negative. It accommodates the growing
understanding that participation in multiple roles can contribute to good mental and physical health so
long as the degree of fit between work and family is satisfactory (Barnett, Garies & Brennan, 1999;
Marks & MacDermid, 1996).

Greenhaus et al. (2003) operationalised the concept of work-family balance as comprising three
components. These are:

Time balance, whereby equal amounts of time are devoted to work and family;

Involvement balance, whereby an equal level of psychological involvement in work and family
roles exists; and

Satisfaction balance, whereby an equal level of satisfaction is derived from work and family
roles.

Within this definition, imbalance occurs when greater time, psychological involvement or satisfaction
occurs in one role compared to the other. Furthermore, Greenhaus et al. (2003) suggest that balance
can be positive where inputs to both roles and outputs (i.e. satisfaction) from both roles are high.
Alternatively, balance can be negative, where inputs and outputs associated with both roles are low.
Work-family balance was conceptualised by Greenhaus et al. as a continuum anchored at one end by
serious imbalance in favour of one role (e.g. work) through a balanced state of equal involvement and
satisfaction with both life domains to another extreme of imbalance favouring the other life domain
(i.e. family). Greenhaus et al. report that individuals who were more engaged in work life than they were
in family life reported higher levels of work-family conflict and stress. In contrast, individuals who were
more engaged in family life than work life reported the highest quality of life and lowest levels of stress.

Work-life or work-family?
The terms work-family and work-life balance have been used in previous studies to describe the extent
to which an individual is equally engaged in work and non-work roles. However, in this study, the term
work-life balance is chosen to broaden the concept to cover individuals who struggle to balance
responsibilities that do not include caring for dependent children, for example, to participate in
sporting, leisure or community activities. This inclusive definition accommodates all employees,
including those who are single and/or child-free, as well as older employees whose children have left
home. The inclusive approach also ameliorates the strongly gendered character of work-life policy
utilisation, in which work-life policies are perceived to be of benefit only to female employees. This is
particularly important in the construction industry where the workforce is predominantly male.
However, in some instances throughout the report the term work-family has been retained where the
original research being referred to has used this term.

The construction industry

The construction industry is a demanding work environment in which a culture of long work hours
prevails. Most construction sites operate on a six-day week, and professionals and managers work many
hours of unpaid overtime. In construction work, job role stressors are also high. Construction
schedules are often tight and severe financial penalties apply if deadlines are not met. Construction

3
industry employees must balance the requirements of multiple stakeholders generating the possibility of
role conflict within the job (Francis & Lingard, 2001). Construction is also a high-risk industry in which
the possibility of safety incidents and serious injury are higher than in other sectors. Construction
managers and professionals thus bear significant responsibility, not only for the completion of projects,
but also for the safety of others. The construction industry traditionally has an adversarial culture in
which conflicts and disputes are commonplace. These characteristics increase the potential for conflict
between employees work and personal lives.

The increase in interest in work-life balance issues has led some construction management researchers
to explore the experience of construction industry employees at the work-family interface. This
research has revealed that site-based employees are particularly vulnerable to high levels of work-family
conflict and burnout. For example, in one Australian construction organisation, the average number of
hours worked each week was 62.5 among site-based respondents in direct construction activity, 56.1
among respondents who work mostly in a site office and 49.0 among respondents in the head or
regional office (Lingard & Francis, in press). Site-based employees also reported significantly higher
work interference with home life and suffered significantly higher levels of emotional exhaustion (a
dimension of burnout) than employees who worked in the head or regional office. A study of
Australian male civil engineers revealed that work-family conflict and work hours were significantly and
positively correlated and that participants did not perceive their organisations to be very supportive of
their needs to balance their work and personal lives (Francis, 2004).

Research has also linked job schedule demands to construction professionals family functioning. For
example, hours worked per week was negatively related to marital satisfaction and positively related to
marital conflict in a survey of Australian civil engineers (Lingard & Sublet 2002). In the same study, the
engineers level of emotional exhaustion (a dimension of burnout) was predicted by their subjective
overload, or a sense of having too much to do in the time available, long work hours and marital
conflict (Lingard, 2003; Lingard, 2004).

The study objectives

The study was undertaken to provide a greater understanding of the work-life balance experiences of
technical and professional construction industry employees in private and public sector organisations.
In particular, the relationship between job and family characteristics, work-life conflict and employee
well-being, family functioning and work-related attitudes were investigated. In addition comparisons
were made between employees experiences in public and private sector organisations. The experiences
of employees in different life stages, for example, child-free employees, employees with pre-school
children, employees with school age children and employees whose children have completed school
were also examined. We tested for significant differences in work hours and other job demands, work-
life conflict, work and family outcomes and individual well-being between groups of respondents, and
examined the bi-variate relationships between variables in the study. Finally, we examined the inter-
relationships between the variables in our study, paying particular attention to the complex ways in
which work and family variables interact to predict work-life conflict and employee burnout.

Conceptual framework

Work-life conflict
Although the work-life balance concept (defined above) recognises that participation in multiple roles
can yield positive benefits to employees, previous research in the construction industry (described
above) indicates that many construction employees suffer from work-life conflict. Work-life conflict is
extensively reported to be a problem for individuals, their families and organisations. In order to
develop effective strategies for preventing work-life conflict it is critically important to gain a good
understanding of its sources and consequences. Work-life conflict was therefore the central variable in
this study.

4
Work-family conflict has been defined as a form of interrole conflict in which role pressures from the
work and family domains are mutually incompatible in some respect (Greenhaus & Beutell, 1985, p.77).
Work-family conflict has been consistently linked to negative outcomes for individuals, families and
employing organisations. For example, work interference with family life has been associated with job
dissatisfaction, life dissatisfaction, turnover intention, general well-being, psychological strain,
psychiatric disorders and substance abuse and problem drinking (Allen et al., 2000; Netemeyer, Boles &
McMurrian, 1996; Boyar, Maertz, Pearson & Keough, 2003; ODriscoll et al., 1992; ODriscoll et al.,
2003; Grant-Vallone & Donaldson, 2001; Hammer, Saksvik, Nytr, Torvatn & Bayazit, 2004; Frone,
2000; Grzywacz & Marks, 2000).

A 2001 study confirmed that not only do Australian construction industry employees suffer from high
levels of work-life conflict, but that work-life conflict is associated with undesirable outcomes including
burnout, diminished organisational commitment and lower levels of marital harmony (Lingard &
Francis, 2002).

Work-life conflict is understood to be bi-directional, for example, family interference with work is
reported to be related to, but distinct from, work interference with family (Frone et al., 1997). Family
interference with work has also been linked to undesirable outcomes, such as job dissatisfaction, life
dissatisfaction, stress and intention to turnover (Kelloway, Gottlieb & Barham, 1999). However,
research has indicated that work demands are much more likely to interfere with family obligations than
the other way around (Frone et al., 1992).

Frone et al. (1997) tested an integrative model of the work-family interface. They discussed the
possibility of domain-specific antecedents and outcomes of work interference with family (WIF) and
family interference with work (FIW). They suggested that work-family conflict acts as a key mediating
variable linking variables in the work domain with variables in the family domain. In particular, Frone et
al. (1997) hypothesised that work distress and work overload gives rise to WIF, which leads to parental
overload and family distress. Conversely, demands in the family environment were hypothesised to give
rise to FIW, leading to work overload and work distress. The results support domain-specific predictors
of WIF and FIW. For example, they found that variables in the work environment, including work
distress and work overload, predicted WIF, while variables in the family domain, including family
distress and parental overload, predicted FIW. The results partially support domain-specific effects in
that FIW was negatively related to work performance and WIF was negatively related to family
performance. This indicates that, if employees obligations in one role frequently interfere with the
enactment of a second role, performance in the second role suffers.

Recently, researchers have identified the lack of a strong conceptual framework as a limitation in much
work-family conflict research (Guerts, Kompier, Roxburgh & Houtman, 2003; Demerouti, Bakker &
Bulters, 2004). These writers point out that most studies have relied on role stress theory which
purports that participation in one role limits the allocation of resources to other roles. However, while
role stress theory provides a good general framework for explaining work-life conflict, it does not
provide an adequate basis for understanding how or why work-life conflict is related to individual,
family and organisational variables.

Guerts, et al. (2003) and Demerouti et al. (2004) point out that, owing to the general nature of role-
stress theory, researchers differ in where they situate work-family conflict in the stressor-strain
relationship. For example, work-family conflict is often regarded as a source of stress which, in addition
to other stressors, will have a deleterious effect on well-being, marital quality or other outcomes. For
example, Mauno and Kinnunen (1999) report that work-family conflict, combined with other stressors
like job insecurity and time pressures at work, predicted marital quality. In other studies, work-family
conflict is conceptualised as an outcome of work and family stressors. For example, workload and work
role conflict have been consistent predictors of work interference with family (Boyar et al., 2003). In
addition to work-family conflict as a source or outcome of stress, other studies report that work-family
conflict mediates the stressor-strain relationship. For example, Guerts et al. report that work
interference with domestic life mediates the relationship between workload and well-being.

5
Guerts et al. (2003) use the Effort-Recovery (E-R) model (Meijman & Mulder, 1998) to explain the
mechanism by which work interference with domestic life mediates the stressor-strain relationship.
Mediator variables explain how or why a predictor variable influences an outcome variable (Baron &
Kenny, 1986). (For further explanation on mediation see section 2, Data analysis). Thus, work-family
conflict is viewed as a mechanism through which work role stress influences well-being. For example,
Guerts et al. (2003) report that work interference with domestic life mediates the relationship between
workload and well-being.

Guerts et al. (2003) suggest that meeting work demands requires effort which produces the tangible
result of the work itself, as well as short-term physiological and psychological reactions in the worker.
These latter reactions are usually reversible after a brief respite from work. However, when workers
have extensive domestic obligations, their opportunity to recover from work impacts is hampered.
Insufficient recovery contributes to higher load reactions with the net negative load effects resulting in
longer term impacts, such as chronic stress, burnout, fatigue and other health problems. This research
adopts this conceptual approach. Work interference with non-work life is conceptualised as a situation
in which recovery in the domains of home, leisure and other non-work domains is impeded by job
demands. Thus, this study hypothesises that WIF mediates the relationship between job demands and
well-being (burnout).

The long-term effects of this interaction are particularly concerning. Demerouti et al. (2004) describe a
situation in which work pressure gives rise to work-home interference and exhaustion. However, in a
three-wave longitudinal study, Demerouti et al. also report reciprocal relationships in which people with
higher levels of emotional exhaustion also experienced more WIF and work pressure over time. Thus
they conclude that common theoretical models postulating the causal chain of work pressure-WFI-
exhaustion are not adequate and that more elaborate models including reciprocal relationships are
needed. Demerouti et al. describe the two-way effect as a loss spiral in which resource loss occurs.
Thus initial losses are follwed by further losses, with each loss resulting in the depletion of resources for
confronting the next threat or loss (Hobfall, 2001). Thus, Demerouti et al. suggest that work pressure
evokes WFI, which consequently brings on feelings of exhaustion. These feelings of chronic fatigue
then lead to greater a perception of work pressure as employees have diminished resources with which
to deal with the pressures of work and higher levels of WFI. The evidence presented by Demerouti et
al. that these loss spirals occur suggests that organisations should take countermeasures to prevent,
rather than to treat, employees experiences of exhaustion and difficulty in trying to manage the work
and home domains. Prevention is particularly important because once these loss spirals have been
entered into, they are likely to be difficult to break.

Antecedents of work-life conflict

Integrative studies of work-life conflict have identified antecedents of work-life conflict in both the
family and work domains (Frone et al., 1997). The most commonly cited antecedents were included as
variables in this study. These are listed below, together with a justification of their inclusion, under the
headings of Organisational antecedents and Family antecedents.

Organisational antecedents of work-life conflict

Job demands
Various job demands have been linked to work-life conflict. Greenhaus and Beutell (1985) identified
two types of direct predictors of work-family conflict. These are time-based and behaviour-based.
Time-based predictors represent role-related time commitments, i.e. the amount of ones time that is
spent involved in work, or family-related activities. Parasuraman, Purohit, Godshalk and Beutell (1996)
suggest that time commitments are an important direct predictor of work-family conflict because time
is a limited resource. Time spent in one role-related activity inevitably reduces the time that can be
devoted to another role. The second type of predictor (strain-based issues) relates to role-related
distress or dissatisfaction (Frone et al., 1997). Greenhaus and Beutell suggested that various role

6
characteristics can generate strain or distress that undermines an individuals ability or willingness to
fulfil the responsibilities of another role. Thus, stressful experiences in one domain may result in
negative mood effects that spill over into another life domain (Williams & Alliger, 1994). Both time-
based and strain-based organisational antecedent variables were included in the current study. These are
described.

Work hours
The construction industry is a demanding work environment in which participants are expected to
work long hours. A recent survey of Australian engineers found that 44% of respondents work 45
hours per week or more, while 21% work over 50 hours. Sixty-seven percent of those working more
than 50 hours per week expressed a preference for working fewer hours (APESMA, 2000). Research
suggests that long work hours are negatively related to family participation and positively related to
divorce rate (Aldous, Osmond & Hicks, 1979). In a study of civil engineers in New South Wales and
Victoria, Lingard and Sublet (2002) reported work hours to be a significant predictor of quality in the
marital relationship. The average number of hours worked per week was positively associated with
marital conflict, and negatively associated with marital satisfaction. The APESMA survey, cited above,
did not examine the impact of hours on relationship satisfaction or conflict but did find that 34% of
respondents indicated some difficulty in taking accrued leave entitlements, and 33% expressed
dissatisfaction on the balance between work and family. Work hours have been consistently linked to
difficulties in balancing work and personal life (Moen & Yu, 2000; Guerts, Rutte & Peeters1999; Batt &
Valcour, 2003). Tausig and Fenwick (2001) reported that as work hours per week increased, employees
degree of work-life balance declined sharply. This makes intuitive sense because time spent at work is
necessarily time that cannot be spent in family, leisure or other activities. Thus, the number of hours
worked each week were included as an antecedent variable in the study.

Weekend work and irregular hours

Construction industry employees are also required to work non-standard work schedules, including
weekend work. Irregularity of work hours has been identified as the most important variable affecting
low marital quality among shift workers (White & Keith, 1990), and non-standard work schedules have
been found to affect separation or divorce rate among married people with children (Presser, 2000).
Although Presser (2000) found no significant relationship between mens weekend work and separation
or divorce rates, regular weekend work emerged from the 2001 pilot study of work and family issues as
a major source of dissatisfaction for project-based Australian construction professionals (Lingard &
Francis, 2002). The results of the pilot study revealed that weekend work was significantly and
positively associated with work-life conflict and emotional exhaustion. Furthermore, dissatisfaction with
work hours and, in particular, weekend work was the most commonly raised theme in qualitative
comments made by respondents in the pilot study. Dissatisfaction with weekend work was most
commonly expressed by project-based employees and those who were partnered with dependent
children (Lingard & Francis, 2002). On the basis of the pilot study results, we determined that the
effects of weekend work were potentially greater than those associated with long work hours between
Monday and Friday. Therefore, weekend work was included as a separate variable in the study.

Subjective workload
Work overload has been a robust predictor of work interference with family (Guerts & Demerouti,
2003; Wallace, 1997). Lingard and Sublet (2002) reported that perceived overload was negatively
correlated with satisfaction in the marital relationship among their sample of practising civil engineers.

Responsibility
One variable not commonly measured in work-life balance research is work-related responsibility. We
included responsibility in the study because construction professionals and managers bear
considerable responsibility for completing projects to tight time schedules and budget constraints.
Failure to meet deadlines result in severe financial penalties, which can threaten the profitability of a
project. Furthermore, increasing safety and environmental imperatives and the threat of personal
liability in the event of unforeseen incidents impose high levels of responsibility in a high-risk industry.

7
For example, Bacharach, Bamberger and Conley (1991) suggest that, for engineers, conflict between
professional standards and budget constraints is strongly associated with severe life and
deathconsequences, increasing the level of responsibility. It is possible that the responsibility borne by
construction industry professionals and managers presents a source of strain-based role conflict
between work and family. Lingard and Sublet (2002) identified some evidence to support the
proposition that work responsibility is a predictor of work interference with personal life. In their study
of civil engineers, they reported that work-related responsibility is negatively associated with the
engineers self-reported satisfaction in their marital relationship. However, work-family conflict was not
measured in the study by Lingard and Sublet (2002), and therefore the mechanism by which work-
related responsibility is linked to family outcomes is unclear. As described above, Frone et al. (1997)
reported that stressors in the work domain affect family functioning indirectly via work interference
with family. Therefore, it is possible that work-related responsibility impacts upon marital satisfaction
via work interference with family. Work-related responsibility was included in our study to test this
proposition.

Family antecedents of work-life conflict

Family responsibilities
While many studies have focused on organisational antecedents to work-family conflict, relatively few
studies have explored the linkage between family responsibilities and work-family conflict. Dual-earner
relationships and the presence of children are thought to greatly increase time demands in the family
domain and increase the potential for work-life conflict. This study examined several family-related
variables as antecedents of work-life conflict. These are described.

Parental responsibilities
The presence of children in a family has been linked to greater work-life balance difficulties. In a study
of technical, professional and managerial employees, Batt and Valcour (2003) reported that both men
and women report that having children in the household lowered their sense of control over managing
work and family. Tausig and Fenwick (2001) report that married couples without children reported
higher levels of work-life balance and that the presence of children whether in single or two-parent
households or dual earner or traditional single earner households is significantly related to lower
balance. Parental demands are believed to be a function of the number and ages of children and the age
of the youngest child (Parasuraman, Greenhaus, & Granrose, 1992). Parental demands are believed to
be greatest for people with infants and pre-school aged children, less for those with school-age children
and lowest for those with adult children who have left home (Parasuraman & Simmers, 2001).
Extensive parental demands require that greater time be committed to family activities which, by
necessity, reduces the time available to the work role.

Parasuraman and Simmers (2001) investigated the relationship between parental responsibilities and
time commitments to family and work-family conflict among self- and organisationally employed
persons. They reported that the pattern of work-family conflict predictors in the family domain differed
according to employment status. Parental demands was positively related to work-family conflict among
self-employed respondents, while family involvement was negatively associated with work-family
conflict among organisationally employed respondents.

Previous research suggests that the presence of children impacts upon work-life balance among
organisationally employed construction professionals and managers. In the pilot study, the number of
dependent children in a household was positively associated with employees self-reported irritability in
parental behaviour and negative spillover between the work and the parenting role. Also, a study by
Lingard (2004) revealed that number of work hours was the only significant predictor of burnout
among child-free Australian civil engineers. By comparison, family variables were the most important
predictors of burnout among engineers who were partnered with children. This suggests that the
development and onset of burnout (which is widely understood to be an outcome of work-family
conflict) is much more complex for employees with dependent children. As a result of this evidence, we

8
included number and age of children as variables in this study to explore the effect of parental demands
on work-life conflict.

Responsibility for other family dependants

Rothausen (1999) is critical of overly simplistic measures of family responsibility. She suggests that the
number of children is particularly problematic because this does not account for the childrens age or
the fact that children of divorced adults may live, for most of the time, with the other parent. Neither
does it accommodate for the added responsibility borne by parents of children with special needs.
Moreover, individuals may have responsibilities to family members other than dependent children, for
example, elderly parents, grandchildren, siblings, children of siblings, spouses or significant others and
in-laws. It is important that the diversity of family responsibilities is not ignored. Indeed, there is
evidence to suggest that caring for different types of family dependants has a different effect on family
life and well-being. For example, caring for an elderly relative in the home is associated with lower
levels of family performance and well-being than caring for a child (Kossek, Colquitt & Noe 2001).
Boyer et al., (2003) define family responsibility broadly as the obligation to care for others who are
either formally or informally sanctioned family members (p.179). Boise and Neal (1996) suggest that
family responsibilities, irrespective of whether these responsibilities involve caring for a child or other
family dependants, increase the time requirements and strain placed on the family. These time
commitments, in turn, can interfere with an individuals work role. However, Boyer et al.explored the
relationship between family responsibilities and family interference with work but, contrary to their
hypothesis, found no significant relationship. Despite this negative result, we included responsibility for
mily dependants other than children in our study so that we would be able to discriminate between the
impact of caring for dependent children and other family responsibilities in our analysis.

Outcomes of work-life conflict

Individual outcomes

Burnout
Burnout is one of the most commonly investigated outcomes of work-life conflict. The most widely
accepted definition of burnout conceptualises the phenomenon as a syndrome of emotional
exhaustion, depersonalisation and reduced personal accomplishment (Maslach, Jackson, & Leiter,
1996). This definition was originally used to describe burnout in the human service or caring
professions but it is now well established that burnout can also occur among other occupational
groups. Thus, Maslach and colleagues have developed a slightly different model of burnout for workers
outside the human services field, in which depersonalisation is replaced by the concept of cynicism.
Cynicism is indicative of a more general withdrawal from work involvement rather than a personal
distancing from an individual client.

In this general burnout model, emotional exhaustion describes feelings of depleted emotional resources
and a lack of energy. In this state, employees feel unable to give of themselves at a psychological
level. Cynicism is characterised by a diminished interest in,and exaggerated distancing from, ones work.
Reduced personal accomplishment refers to a situation in which employees tend to evaluate themselves
negatively and become dissatisfied with their accomplishments at work.

Research suggests that burnout is associated with negative outcomes for both individuals and
organisations. At an individual level, burnout has been associated with the experience of psychological
distress, anxiety, depression, reduced self-esteem and substance abuse (Maslach et al., 2001). Some
studies also identify a link between burnout and coronary heart disease (Appels & Schouten, 1991;
Tennant, 1996), suggesting that burnout should be treated as part of a companys occupational health
progam. Research also consistently links burnout to lower levels of organisational effectiveness, job
satisfaction and organisational commitment as well as to higher levels of absenteeism and turnover
(Wright & Bonett 1997; Maslach et al., 1996; Schaufeli & Enzmann, 1998). In a recent study of
Australian civil engineers, the emotional exhaustion and cynicism dimensions of burnout were strong
predictors of engineers intentions to quit their jobs (Lingard, 2003).

9
Burnout was originally believed to be a uniquely work-related phenomenon. However, it is now
acknowledged that burnout is the result of a complex interaction of factors in employees work and
non-work lives (Maslach et al., 2001; Carroll & White, 1982). Empirical research data support this
contention, particularly in the case of dual earner couples (Dolan, 1995; Gmelch & Gates, 1998;
Lingard 2004). We therefore included burnout in our study as a possible outcome of work-life conflict.

Family outcome variables

Marital satisfaction
Several studies have linked aspects of the work environment to quality in employees relationships with
their spouses, partners or significant others. For the purposes of brevity, we will refer to all of these
relationships as marital relations. For example, Mauno and Kinnunen (1999) reported that job demands
were indirectly related to marital satisfaction and that this relationship is mediated by individual well-
being. Crouter, Bumpus, Head and McHale (2001) investigated the extent to which work hours and
overload negatively affect mens family relationships. They reported that mens long work hours were
unrelated to marital quality (measured by marital love and marital conflict). However, all three
dimensions of marital quality were affected by mens feelings of subjective overload (see subjective
workload above). Thus, although men who work longer hours (more than 60 hours per week) spent
less time with their wives, this did not reduce both partners subjective assessments of the quality of the
marital relationship. However, when men felt overloaded, their partners felt less loving, saw themselves
as less able to take their spouses perspective, saw their spouse as being less able to take their
perspective and reported more marital conflict. Crouter et al. (2001) suggest that work hours do not
have a negative impact on the quality of mens marital relationship because in many partnerships both
husbands and wives allocate to men the primary breadwinning role. Thus, mens time commitments to
work are regarded as necessary in order to maintain the familys preferred lifestyle, for example
payment of school fees, mortgages etc. In contrast to this finding, Lingard and Sublet (2002) reported
that work hours per week was a significant predictor of employees perceived quality of their marital
relations in a sample of Australian civil engineers. In this latter sample of predominantly male civil
engineers longer work hours was associated with lower levels of relationship satisfaction and higher
levels of marital conflict.

However, in the study by Lingard and Sublet (2002), the mechanism by which work hours predicted
marital quality was not investigated. In accordance with the integrative model of work and family
interface proposed by Frone et al. (1997) demands in the work domain should affect family via work
interference with family. Thus, marital relationship quality was included as a variable in our study. It was
hypothesised that work interference with family would mediate the relationship between job demands,
including work hours, and marital quality.

Parent-child relationship
Previous research has also linked parental experiences at work to consequences for their children. For
example, Stewart and Barling (1996) suggest that parents work experiences indirectly influence
childrens behaviour through parents work-related affect and parenting (p.222). Crouter, Bumpus,
Maguire and McHale (1999) empirically tested this hypothesis among a sample of dual earner families
with two adolescent offspring. They reported parents work pressure impacts upon adolescents well-
being through parents feeling of being overloaded and conflict in the relationship between the parent
and the adolescent. In one study by the Australian Institute of Family Studies, children were asked to
comment on their experience of parental work patterns in their family. Primary school-aged children
indicated that it was particularly important to them for their parents to be involved in aspects of their
lives such as attending school events and picking them up from school (Lewis et al., 2001). If work
schedule demands prohibit this involvement, it is possible that parent-child relationships could suffer.

In the 2001 pilot study, parents self-reported irritability in their relationship with their child(ren) was
positively associated with their perception that work interfered with their parenting responsibilities and
leisure time. Conversely, parents self-reported affection in their relationship with their children was

10
negatively associated with their perception that work interfered with their parenting responsibilities and
leisure time. The perception of work spillover into parenting and leisure time was very strongly and
positively associated with the number of hours worked each week (Lingard & Francis 2002). Thus,
when it is perceived that work negatively interferes with parenting and leisure, parents who work in the
Australian construction industry report lower levels of quality in their relationships with their children.
For these reasons, parents perceptions of the quality in the parent-child relationship were measured in
the study. In accordance with the model proposed by Frone et al. (1997), it was hypothesised that work
interference with family would mediate the relationship between job demands, including work hours,
and quality in the parent-child relationship.

Organisational outcome variables

Organisational commitment
Organisational commitment has been predicted by organisational supportiveness of employees family
lives in previous studies (Orthner & Pittman, 1986; Grover & Crooker, 1995).
Organisational commitment research has identified three facets of commitment. These are:

Affective commitment which reflects an emotional attachment to the organisation, whereby
employees remain in the employment of an organisation because they want to;

Continuance commitment which reflects employee motivation to remain with an
organisation because they feel they have to; and

Normative commitment, which reflects an employees sense of moral obligation to stay with an
organisation (Allen & Meyer, 1990)

Affective commitment is the dimension of commitment most strongly and consistently associated with
employee performance and organisation-serving behaviours such as effective use of time. In contrast,
continuance commitment has been negatively linked to employee performance and positively linked to
voluntary absence (Meyer, Allen & Smith, 1993). Meyer, Paunonen, Gellatly, Goffin and Jackson,
(1989) concluded that the value of commitment to an organisation depended upon the form that
commitment took.

Several studies have found a negative relationship between work interference with family (WIF) and
affective commitment (Thompson, Beauvais & Lyness, 1999; Netemeyer et al., 1996). By comparison,
studies of work-family conflict and continuance or normative commitment are few. However, Casper,
Martin, Buffardi and Erdwins (2002) investigated the impact of both work interference with family
(WIF) and family interference with work (FIW) on affective and continuance commitment in a sample
of employed mothers. They hypothesised that there would be a negative relationship between WIF and
affective commitment because WIF is associated with undesirable stress-related outcomes. Since the
sources of WIF reside in the work domain, it is expected that individuals high in WIF would feel less
emotionally attached to their organisation. However, contrary to Casper et al.s hypothesis, they report
no negative relationship between WIF and affective commitment. This is not in line with the results of
previous findings, and Casper et al. suggest that this may be because, unlike other studies, the sample
was entirely female. They suggest that differences may exist between the ways in which men and
women, and parents and childless employees, manage the work-family interface. For example, they
suggest it is possible that employed mothers of young children use segmentation to manage work-
family issues and, therefore, that WIF may not influence affective commitment as it would for fathers
or child-free women. Owing to the lack of consistent results in previous research, and the fact that our
sample was likely to differ from Casper et al.s by being predominantly male, we included affective
commitment as a variable in our study and explored the relationship between WIF and affective
organisational commitment.

Casper et al. (2002) also hypothesised that there would be a positive relationship between WIF and
continuance commitment on the basis that individuals whose jobs result in them experiencing WIF
were likely to experience conflict and guilt. They suggested that in order to reconcile these feelings,
employees attribute the reason for them staying in jobs which cause WIF as necessity, i.e. they stay
because they have to. Their results supported this hypothesis, suggesting that employees who
experience WIF but remain employed by their organisations, attribute their continued employment to

11
the need to work (continuance commitment) to avoid feelings of guilt that arise as a result of their
perception that their work interferes with their family. We explored the relationship between
respondents WIF and continuance commitment.

The relationship between family interference with work (FIW) and organisational commitment is also
inconclusive. Some studies have found a negative relationship between family interference with work
(FIW) and affective commitment, while others have found no significant relationship (ODriscoll et al.,
1992). Consequently, we explored the relationship between FIW and both the affective and
continuance dimensions of organisational commitment.

Job satisfaction
One of the most commonly investigated correlates of work-family conflict is job satisfaction. Job
satisfaction was measured in our study because work-family conflict has been linked to general job
satisfaction, satisfaction with supervision and work load (Boles, Howards & Donfrio, 2001). In most
studies, job satisfaction is reported to decrease as work-life conflict increases; however, reviews of the
literature on work-family conflict and job satisfaction have identified inconsistent results (Kossek &
Ozeki, 1998). Bruck, Allen, and Spector (2002) conducted a finer grained analysis of the relationship
between work-family conflict and job satisfaction. They used a bi-directional measure of work-family
conflict, which measured both family interference with work (FIW) and work interference with family
(WIF). They also used a composite or facet-specific job satisfaction measure, which measures the
attitudes a person has with regard to various aspects of their job: for example, co-workers, pay and
opportunities to learn new things. Bruck et al. suggested that individuals often differed in their degree
of satisfaction across facets, for example, an employee might be satisfied with their pay but dissatisfied
with their job security.

Bruck et al (2002) hypothesised that composite job satisfaction would correlate more strongly with
work-family conflict than global job satisfaction, because work-family conflict is likely to arise as
specific aspects of work make it difficult to balance work and family demands. Bruck et al. empirically
tested their hypotheses, and their results confirm that composite job satisfaction is more strongly
correlated with work-family conflict than was global job satisfaction.

Bruck et al. (2002) also explored the effects of both FIW and WIF on job satisfaction. They report that
FIW did not add unique variance beyond that of WIF in explaining variance in job satisfaction. This
means that, after accounting for the extent to which work interferes with family, family interference
with work has little relationship with job satisfaction. This indicates the existence of domain-specific
relationships in which conflict originating from one domain (in this case, work) has a stronger
relationship with work-related attitudes and behaviours than conflict arising from another domain (for
example, family).

Turnover intention
Greenhaus, Parasuraman and Collins (2001) suggest that, when employees experience work-family
conflict, they will try to eliminate this conflict by withdrawing from work. In particular, Greenhaus et al.
report that employees who are dissatisfied with work and family think about quitting. Higher levels of
stress associated with work-family conflict are also linked to withdrawal tendencies. In their study of
technical, professional and managerial employees, Batt and Valcour (2003) reported work interference
with family to be significantly and positively related to turnover intentions, and employees perceptions
of control over managing work and family to be significantly negatively related to turnover intention.
We therefore included turnover intention as an outcome variable in our study. Netemeyer et al. (1996)
examined both family interference with work and work interference with family, and found both to be
correlated with turnover intention in two out of three samples. In a more recent study, Boyar et al.
(2003) also found that both directions of work-family conflict predicted turnover intention. They
concluded that reducing work-family conflict could help to reduce costly employee turnover.

Labour turnover involves redundant monetary and non-monetary costs, which can have a detrimental
impact upon organisational effectiveness. This is particularly true when experienced or competent

12
workers with substantial investments in their human capital (e.g. education, professional accreditation,
skills etc) leave an organisation voluntarily. Like Netemeyer et al. (1986), Boyar et al, (2003) and Batt
and Valcour (2003), we did not measure actual turnover behaviour. However, research suggests that
intention to turnover is a good predictor of actual turnover behaviour (Parasuraman, 1982). Hughes
(2001) also notes that intention to leave may be even more damaging than actual turnover, because
such intentions can be constrained by the availability of acceptable alternatives. Where such alternatives
do not exist, employees may involuntarily remain in their job, which poses additional possible
performance impacts (Hughes, 2001).

Moderator variables

Several variables were included in our study as moderator variables. Moderator variables are variables
that affect the direction and/or strength of the relationship between an independent/predictor variable
and a dependent variable (Baron & Kenny, 1986). For further information on moderation, refer Section
2, Data analysis. Moderator variables measured in this study are described below.

Organisational and social support

Organisational and social support (from supervisors and co-workers) has been linked to employees
willingness to use formal work-life balance policies. Recent studies have demonstrated that, in order for
employees to experience reductions in work-family conflict, in addition to formal policies, there is a
need to ensure that they work in a supportive work environment (Thompson et al., 1999; Allen, 2001;
ODriscoll et al., 2003). A recent study of Australian civil engineers revealed that engineers who
reported their organisations to be supportive of their family commitments also reported a higher levels
of organisational commitment and job satisfaction as well as lower intention to quit (Francis, 2003). In
contrast, unsupportive work cultures in which employees perceive negative career consequences
associated with work-family policy usage, in which managers are not perceived to be sensitive to
employees family responsibilities and in which the culture equates long work hours with employee
commitment are likely to impede the use of work-family balance initiatives, undermining their
effectiveness.

Evidence exists to suggest that the Australian construction industry culture is not supportive of
employees work-family balance. For example, Lingard and Lin (2003) conducted a survey of women
working in the construction industry of Victoria and reported that although part-time work options and
parental leave entitlements were available to the majority of respondents, many women indicated they
were unlikely to use these entitlements while working with their current employer. Half of the
respondents indicated that they thought that taking maternity leave or utilising part-time work options
would have a negative impact on their career.

In the general organisational and social psychological literature, a hypothesis often examined is the
extent to which social support has a buffering effect, mitigating the harmful effects of stressors in
stressor-strain relationships. In the work-family literature, some studies have confirmed that the
presence of social support reduces the negative consequences of work-related stressors and work-family
conflict (Thomas & Ganster, 1995; Goff, Mount & Jamison, 1990). However, other studies have
revealed no such effect. Despite these inconsistent results, it seems feasible that organisational,
supervisory or co-worker support could help protect individuals from experiencing the negative effects
of work-life conflict.

Organisational and social support variables were therefore measured and tested for their moderation
effects in our study.

In particular, we examined:
1. The extent to which perceived organisational support moderated the relationship between
work-life conflict and organisational commitment and job satisfaction; and
2. The extent to which supervisor support and co-worker support moderated the relationship
between work-life conflict and employees psychological strain (burnout).

13
It was anticipated that, when supervisor or co-worker support were high, work-life conflict would be
more weakly and positively related to employee burnout than when supervisor or co-worker support
was low. It was also expected that when perceived organisational support was low, work-life conflict
would be more strongly and negatively related to organisational commitment and job satisfaction than
when perceived organisational support was high. The rationale for including the organisational and
social support variables as moderators is presented below.

Perceived Organisational Support (POS)

Eisenberger, Huntington, Hutchison and Sowa (1986) suggest that individuals develop global beliefs
concerning the extent to which the organisation values their contribution and cares about their well-
being. They termed these beliefs Perceived Organisational Support (POS). POS has been found to
influence employee diligence, commitment and innovation (Eisenberger et al., 1990). Allen (2001)
reports that perceptions of organisational support of work-family balance are associated with reduced
work-family conflict and enhanced job satisfaction and organisational commitment.

Casper et al. (2002) explored the extent to which POS can act as a buffer, preventing work-family
conflict from adversely impacting upon organisational commitment. The rationale for this hypothesis is
that employees who work for supportive organisations are likely to experience reduced stress as a result
of work-family conflict resulting in greater emotional attachment (affective commitment) to the
organisation, and a greater sense of loss associated with leaving an organisation. In a sample of
employed women with children, Casper et al. reported that when WIF is high, and where FIW and
POS were also high, continuance commitment was reduced, suggesting that women experiencing bi-
directional work-family conflict and a supportive environment were less likely to feel trapped in their
jobs.

Given the increasing interest in the effect of organisational support on employees experiences at the
work-family interface, and lack of investigation of POS in the construction industry, we included POS
as a moderating variable in our study.

Supervisor support
Having a supportive supervisor has been reported to reduce the negative impacts of work-family
conflict (Thomas & Ganster, 1995). Some studies have indicated that supervisor supportiveness buffers
the effect of work-family conflict on employees psychological strain. For example, ODriscoll et al.
(2003) tested whether supervisor support for employees work-family balance moderated the
relationship between work-family conflict and psychological strain. They report that, when work-to-
family interference is high, respondents who reported high levels of supervisor supportiveness
experienced less psychological strain than those who reported low levels of supervisor supportiveness.

Supervisors play a particularly important role because they often act as gatekeepers in controlling access
to alternative work arrangements (Watkins, 1995). For example, Barham, Gottlieb and Kelloway (2001)
report that supervisor/manager responses to employee requests for alternative work schedules are
influenced by the employee gender, the nature of the employees care-giving role and employee work
responsibilities. Supervisors were less likely to grant alternative work arrangements that reduced
employees work hours where requests were made by men, where requests were to provide care for an
aged relative rather than a child and when the employee requesting the alternative work arrangement
was a manager.

Supervisors can buffer the effects of work-life conflict caused by organisational demands. For
example, in cases where employers view the use of work-family policies as being indicative of low levels
of commitment, supervisors could mitigate this unsupportive organisational norm by demonstrating
that utilisation of work-family benefits will not be held against employees in the workgroups they
supervise. Supervisors can also suggest informal ways to assist employees who experience difficulty in
balancing work and family (Batt & Valcour, 2003). In other cases, supervisors could discourage the use
of flexible work practices in their work groups believing that this could impact negatively upon work

14
group productivity. Thus, we hypothesised that when supervisory support for employees work-life
balance was high, the relationship between work-life conflict and psychological strain (emotional
exhaustion) would be weaker than when supervisory support was low.

Co-worker support
Research also suggests that social relations among co-workers impact upon employees experiences at
the work-family interface. To some employees, the idea of assisting other employees to balance work
and non-work lives may appear unfair. For example, taking time off work to care for a sick child may
result in an increased or perceived burden on the other remaining employees. Resentment can also arise
when co-workers with no immediate family responsibilities perceive that benefits enjoyed by other
employees are not available to them (Hegtvedt, Clay-Warner & Berrigno, 2002). For example, single or
child-free employees may feel excluded from certain benefits, such as child-care or dependent care
support. Blair-Loy and Wharton (2002) report that, among a sample of professional and managers,
work-group characteristics had a modest influence on employees use of family-care policies, but a
powerful effect on their use of flexibility policies. They suggest that flexibility policies require more
schedule adjustment on the part of co-workers.

Kirby and Krone (2002) suggest that a situation can arise in which work group members do not want
others to make use of flexible work arrangements, even when these are formally available as
organisational policies. They reported that employees in work groups created their own cultural
norms about family-friendly policy utilisation and that work-group members conversations about
policy utilisation are reported to shape group members beliefs about what constitutes use and abuse
of such policies. It is likely that these work group norms influence employees behaviour in balancing
work and family. In particular, when co-workers are not supportive of others attempts to balance work
and personal lives, work-life conflict is likely to result in heightened psychological strain.

In the construction industry, work is largely performed in teams and it is therefore likely that, in this
context, co-worker support concerning employee family responsibilities will impact upon employees
experiences. Furthermore, Hegtvedt et al. (2002) report that resentment of the need to perform extra
work when other employees make use of work-life balance initiatives is high among men and
employees with traditional gender-role attitudes. In the male-dominated construction environment, co-
worker support for employee efforts to balance work and personal life might be a particularly important
variable moderating the relationship between work-life conflict and psychological strain. Thus, we
hypothesised that when co-worker support for employee work-life balance was high, the relationship
between work-life conflict and psychological strain (emotional exhaustion) would be weaker than when
co-worker support was low.

Flexible work arrangements

Barnett at al. (1999) refute the widely held view that work-family conflict is inevitable when people
simultaneously occupy two or more roles. Indeed, they suggest occupying multiple roles can contribute
to good mental and physical health. However, in order for these positive benefits to be enjoyed, work
arrangements must be suited to family needs. Fit refers to the extent to which work arrangements
meet family needs. Barnett et al. (1999) suggest that when fit is good, conflict will be low, but when fit
is poor, conflict will be high.

Flexible work arrangements can increase the level of fit between employees work and family demands.
Research suggests that flexibility - or employees ability to exercise a measure of control over when and
where they work - can assist them to balance long work hours with family and personal life.

Flexibility in work arrangements can enable employees to integrate and overlap their work and family
responsibilities, and assist them to achieve a healthy balance between work and family. For example,
flexible hours give employees a degree of control over when they choose to fulfill their work
requirements while flexplace (flexible place of work) policies give employees some control over where
they perform their work. Flexplace includes telecommuting, which is the option for employees to work
from another fixed location off-site, typically their home.

15
Flexible work arrangements can yield personal and organisational benefits. For example, if work
arrangements are rigid, an employee with a dependent child may take the day off to take the child to the
doctor. With flexible work arrangements, the employee can take the sick child to the doctor, return
home and continue to work, albeit at a slower pace while caring for the sick child.

Hill, Hawkins, Ferris and Weitzman, (2001) investigated the extent to which flexible hours and
flexplace arrangements resulted in benefits for individuals (more satisfactory work-family balance) and
organisations (longer work hours). They reported that, given a reasonable work week (which they
considered to be between 40 and 50 hours per week), employees who perceived flexibility in the timing
and location of their work had less difficulty with work-family balance. Moreover, Hill et al. reported
that employees with perceived flexibility in the timing and location of work could work longer hours
before work-family balance was perceived to be a problem. From a business perspective, Hill et al.s
results suggested that, given a heavy workload, perceived flexibility in the timing of work enables
employees to work an extra day a week i.e. 60 hours a week, compared to 52 hours without flexible
hours.

Tausig and Fenwick (2001) also explored the effect of the availability of alternate work arrangements
and employees perceived level of control over their work schedule. They found that the mere
availability of alternate work schedules, for example, flexible hours, extended lunch breaks etc, per se did
not facilitate employee work-life balance. What mattered more, according to Tausig and Fenwick, is
employee perceptions of control over their schedule. This perception of control was positively related
to better work-life balance in all groups of employees, including all of six different configurations of
family. This suggests that work-life balance can be managed where employees are given the ability to
select work schedules that balance their work and personal obligations. This finding is echoed in
research by Eaton (2003), who reports that the presence of formal and informal work-family policies
was unrelated to organisational commitment, but that employees perceived control over their working
time, flexibility and pace of work directly predicted both organisational commitment and productivity
outcomes in a sample of male and female technical, managerial and professional employees.

ODriscoll et al. (2003) suggest that the extent to which individuals have control and flexibility is a
critical factor in the reduction of work-family conflict and its impact on well-being. Thus, in this study
we measured respondents flexibility and control over their work arrangements and tested the extent to
which this flexibility and control moderated the relationship between work schedule demands and
work-family conflict. We expected that flexibility and perceived control would both have a buffering
effect, such that when an individuals sense of control was high the relationship between job schedule
demands and work-family conflict would be weak. However, we expected that when flexibility and an
individuals sense of control were low, job schedule demands would be strongly and positively related
to work-family conflict.

Work-life benefits

Many studies suggest so-called family-friendly benefits provide a competitive edge to employers in
attracting and retaining high-calibre employees. Indeed, a study by Grover and Crooker (1995) suggests
that work-family benefits have a positive impact upon employees attachment to their organisation,
irrespective of whether they actually use these benefits or not.

However, Bardoel, Moss, Smyrnios and Tharenou (1999) suggest that, in order to effectively recruit and
retain employees, work-life policies and programs must be strategically designed to meet the needs of
key constituent employee groups. Blair-Loy and Wharton (2002) support this argument reporting that
even in a homogeneous sample of managers and professionals, family-care and flexibility policies were
mostly used by different employees with different individual characteristics. It was therefore valuable, in
this study, to explore the types of work-life balance initiatives which employees in the two participating
organisations would like access to.

16
In a similar manner, McKeen and Burke (1994) explored the work-life initiatives valued by managerial
women and found significant differences according to age and parental status. Secret (2000) also
investigated the effect of individual and job-related variables on employees use of work-family benefits.
Secret (2000) classifies work-family benefits into four categories:

Alternate work arrangement benefits;

Paid leave time benefits;

Child care services benefits; and

Mental health benefits.

Alternative work arrangements include the modification of daily start and stop times, compressed work
weeks, part-time work, job-sharing and telecommuting on a regular basis. Leave time policies include
mandated maternity and paternity leave entitlements and informal arrangements for a few hours or days
off with or without pay but available on an ad hoc basis. Mental health and wellness services include
Employee Assistance Programs, stress management workshops and seminars on family-related matters.
Dependent-care services include on-site childcare facilities, subsidization of child care costs, pre-tax
credits for child-care assistance and information referral services for dependent care and resources.

Perhaps, surprisingly, Secret (2000) reported that employees with dependent children were no more
likely than other groups to use any benefits, suggesting that work-family benefits, especially alternative
work schedules, are valued by all employees, including those who are child free. The implication of this
finding is that such strategies can be useful to employers who are concerned about equitable provision
of benefits to employees with and without dependent children. Also, in Secrets study, women were
more likely than men to use paid leave and mental health benefits only when they experience a family
crises and child care problems respectively. Secrets finding that gender is not a particularly strong
predictor of use of work-family benefits makes it important and interesting to explore the extent to
which employees value work-life balance initiatives in the male-dominated realm of the construction
industry.

Public/private sector differences

Our study was unusual in the fact that both public and private sector organisations were represented.
This enabled a comparison to be made between the work-life experiences of technical, professional and
managerial employees in both public and private sector construction organisations. Research suggests
that public sector employees experience less conflict between their work and personal lives than
employees in private sector organisations (Tausig & Fenwick, 2001). Tausig and Fenwick suggest that
this may be due to the regularity of schedule in most public sector jobs. This regularity, they suggest,
permits employees to plan for their personal lives, as well as helping to differentiate work from personal
or family time. It is also argued that public sector organisations have led the way in promoting work-
family initiatives (Kamerman & Kahn, 1987). Ingram and Simons (1995) report public sector
organisations to be more responsive to work-family needs than private sector organisations. They
explain this by suggesting that work-family issues are more congruent with the organisational goals and
policies of public sector organisations than private sector organisations, because public sector
organisations are responsive to social welfare concerns. They argue that in the public sector, the goal is
to deliver social services and organisations and their members are not held exclusively to economic
standards of performance.

Research also suggests that employees in the public sector might be more likely to use organisational
work-life balance initiatives than their counterparts in the private sector. Secret (2000) reports that
workplace structural factors are better predictors of use of work-family benefits than individual
employee attributes. In particular, employees in larger organisations and those in supportive work
cultures are more likely to use paid leave entitlements than those in smaller organisations or less
supportive cultures. Secret (2000) reported that employees in non-profit or public sector jobs were
more than twice as likely to take advantage of alternative work arrangements, and 1.18 times more likely
to take paid leave entitlements, than those in the private sector.

17
Ingram and Simons (1995) suggest that large organisations are likely to be more subject to institutional
pressure and the need for legitimacy than smaller organisations, due to the fact that large organisations
receive more attention from the public, the media and regulators. Also, the legitimacy and norms and
practices within an organisational field or industry are likely to influence an organisations
responsiveness to work-life issues. Ingram and Simons (1995) report that, where institutional pressures
are high and where other organisations have implemented family-responsive policies, employers are
more likely to implement work-life balance initiatives. It is possible that institutional determinants of
policies supportive of work-life balance are stronger in public sector organisations than in the private
sector. In-keeping with previous studies, we expected that work-life conflict would be lower among
public sector respondents to our survey than among private sector respondents.

Domestic partners

Our study was also unusual in that we invited the domestic partners of employees to complete a shorter
version of the survey. Domestic partners were only invited to participate in the study if employees
agreed to their participation. The rationale for inviting employees domestic partners to participate in
the study is explained below.

Many studies of work-family conflict have focused on the individual as a unit of analysis, identifying
within-individual determinants and outcomes of work-family conflict. This approach assumes that the
work and family role behaviour of an individual is independent of the work and family role behaviour
of his or her domestic partner. It is now recognised that this assumption is not valid and that there is
benefit in studying the couple as the unit of analysis because this permits research to examine how the
stress or strain effects can be transmitted between partners. In seeking to measure the experiences of
both employees and their domestic partners, we intended to examine the crossover effects of work and
family variables within couples.

The effect of partners career/job involvement

Gupta and Jenkins (1985) comment on the fact that, in dual-earner couples, in addition to the pressures
posed by employees own work roles, work-life balance may be influenced by the interaction of their
own role with those of their partners. The reasoning for this is that an employee whose partner is highly
involved with, and committed to, their job is likely to experience greater family pressures because their
partner is likely to devote considerable time and energy to his/her own work role and will have less
time to participate in family life. There is some evidence to suggest that men married to career-
committed women experience higher levels of work-family conflict than men whose wives are less
career committed. For example, Greenhaus and Kopelman (1981) reported that men experienced
higher levels of work-family conflict when their wives held managerial or professional jobs.

Greenhaus, Parasuraman, Granrose, Rabinowitz and Beutell (1989) further investigated the interaction
between partners work salience and employees work-family conflict. They collected data from male
and female partners in dual-career couples. For the purposes of their study, career was defined broadly
as people engaged in a pattern of work-related experiences that span the course of a persons life
(Greenhaus et al., p.138). Greenhaus et al. reported that womens work-family conflict was predicted by
their work-salience, suggesting that when women were highly involved in their own work, they could
devote less time and energy to family life and experience higher levels of work-family conflict as a
result.

Interaction effects were not significant for women in the sample, suggesting that male partners work
salience does not impact upon womens work-family conflict. In contrast, mens own work-salience was
unrelated to their work-family conflict, but the interaction between mens work salience and that of
their partners did predict mens work-family conflict. When men and women both placed high priority
on their careers, men experienced higher levels of strain-based work-family conflict. Interestingly,
Greenhaus et al. (1989) also report that men who displayed high levels of job involvement, and whose
partners also displayed high levels of job involvement (as opposed to career priority), experienced
relatively low levels of time-based work-family conflict. Greenhaus et al. suggest this may be because

18
when partners are both highly involved in their jobs, they better understand one anothers needs, can
provide support to each other and develop a mutually accommodating relationship.

In our study we examined the extent to which partners time commitments to paid work, job
involvement and perception of career priority (relative to the employees career) interacted with
structural characteristics of employees own jobs (for example time demands, levels of responsibility
and inflexibility) to predict employees work-life conflict. We hypothesised that employees whose
partners devoted relatively more hours to paid work each week, were relatively more involved in their
jobs and who perceived their own career to be of equal or greater importance to that of the employees,
would experience higher levels of work-life conflict.

Crossover effects
Crossover effects describe how stress and/or strain experienced by one partner affects the role
conflict, family functioning or well-being of the other partner. The crossover of work-life experiences
from one partner to another has also been investigated with regard to a number of dependent variables
including marital functioning (Mauno & Kinnunen, 1999), work-family conflict (Hammer et al., 1995)
and burnout (Westman et al., 2001). We sought to replicate some of these statistical tests.

In a sample of dual-earner couples, Hammer et al. (1995) explored the extent to which female and male
experiences of work-family conflict had a significant impact upon their partners levels of work-family
conflict. Their results revealed that male work-family conflict had a significant effect on female work-
family conflict and also that female work-family conflict had significant effects on male work-family
conflict. This strong crossover effect indicates that an individuals level of work-family conflict is a
significant predictor of their partners work-family conflict. Furthermore, Hammer et al. (1995) report
that, for males, crossover effects from female partners work and family variables accounted for almost
as much variance in work-family conflict as did within-individual determinants (11 and 12% of variance
respectively). For females, crossover effects from their partners work and family variables accounted
for 7% of the variance in work-family conflict compared to 17% explained by within-individual
variables. This suggests that men may be more susceptible to the transmission of work-family conflict
from female partners than women are from their male partners.

Hammer et al. (1995) conclude that adopting a family systems approach can provide a better
understanding of the sources of work-family conflict. Following Hammer et al. (1995), we explored the
extent to which there was a crossover effect between work-family conflict experienced by employees
and their partners in our sample of matched partners. In particular, we examined whether partners
work-family conflict explained unique variance in employees work-family conflict after employees
within-individual work and family predictors of work-family conflict had been accounted for.

Westman and Etzion (1995) explored the transmission of burnout in a sample of 101 male military
officers and their wives. They reported that burnout had significant crossover effects from one spouse
to the other. They suggest that burnout in one partner creates an additional source of stress, which
leads to burnout in the other partner.

Finally, Mauno and Kinnunen (1999) tested the hypothesis that job stressors are directly related to
emotional exhaustion and indirectly related to the level of marital satisfaction enjoyed by the individual.
They also hypothesised that there would be a crossover effect from an employees level of emotional
exhaustion to his or her partners marital satisfaction. Thus, when employees emotional exhaustion was
high, they suggested that there would be diminished marital satisfaction in both the employee and his or
her partner. However, Mauno and Kinnunen (1999) found no empirical support for the crossover
process between partners. While individual emotional exhaustion mediated the relationship between job
demands and marital satisfaction, emotional exhaustion experienced by one partner did not cross over
to affect the other partners marital satisfaction.

19
20
Section 2: Methodology
Data collection and sampling

The data were collected from employees of a public sector and private sector organisation in
Queensland, Australia. The public sector organisation is involved in the construction of large
infrastructure projects, and the private sector construction company is involved in both civil
engineering and building projects. Project based managerial staff, and their domestic partners, were
specifically selected to participate in the study.

The majority of the data was collected via internet based surveys (see below for further information on
internet surveys). A website was established for the study at www.ciiastudy.com.au and two different
surveys were available for employees and their domestic partners. Paper-based surveys were made
available on the website (downloadable PDF version) and within each organisation to allow for
respondents who were not comfortable completing the survey on-line to participate in the study. The
website contained information on the study, as well as details on how confidentiality would be
maintained. Each respondent, on starting the survey automatically generated a code to enable matching
of domestic couples for data analysis.

A letter from senior management explaining the purpose of the survey and assuring confidentiality and
anonymity of responses was sent to each potential respondent of the employee survey. Within this
letter employees were encouraged to invite their spouse/partner to be part of the study. Details of the
website address, employee and partner logins and passwords were also provided. Three weeks after this
initial letter a follow-up e-mail was sent to all participants reminding them to submit their
questionnaires, if they wished. Members of the steering committee, working for the private and public
sector organisations estimated the sample size to be 110 and 500, respectively and anticipated half of
the sample to be partnered.

Limiting a survey to a single private and public organisation can pose problems for the generalisability
of its findings. However, there are also advantages in that context variables such as organisational
culture, policies and procedures are controlled. In the light of the scope and purpose of this study, data
collection from one organisation was deemed to be adequate.

Internet survey
Internet surveys are increasingly used by organisational behaviour researchers and consultants. Internet
surveys have been found to result in more and longer write-in comments on open-ended survey items.
Internet surveys also enable quicker data analysis and, arguably allow for faster implementation of
policies based on the data collected. Thompson, et al. (2003) report employees positive response to a
web-based organisational climate survey, which yielded a higher response rate than a paper-based
equivalent. Furthermore, there was no indication that the web-based survey discouraged the
participation of any one type of employee (based on gender, race, employee category and work group
size). They reported that demographic sub-group comparisons revealed no significant differences in
response rates or attitudes towards on-line surveys, suggesting that the web-based survey medium does
not tend to over sample any particular demographic sub-group. Stanton (1998) administered an
organisational fairness and supervisory consistency survey to two samples using an internet method and
a paper-based method and reports that the internet data were more complete than data collected using
paper and pencil methods. Thus the problem of handling missing values was minimised. Internet data
also showed comparable variability to data collected using traditional survey techniques, indicating that
respondents were not more likely to check the same box throughout an internet survey without
thinking about the question, than they were in a paper-based survey. The psychometric properties of
scales were also found to be acceptable when the scales were administered in an internet survey. For
example, the factor structures of multi-dimensional scales did not differ across the two data collection

21
formats and the assignment of items to factors was also identical. The relationships between variables
were also equivalent across the two datasets. These results suggest that it is appropriate to use the
internet to administer existing scales in organisational surveys. Internet surveys have been found to
yield high response rates in previous work-life balance studies (see example, Hill, et al. (2001)).

One disadvantage of using the internet as a survey delivery medium is that managerial and professional
people will be overrepresented among respondents. However, as the target audience for our survey was
professional, managerial and technical employees, this limitation was not believed to pose a problem. A
second potential problem associated with freely available internet surveys is that response rates are
impossible to calculate and the researcher has no control over who may complete the survey. Again,
this was not a problem in our study because the internet survey utilised an access control system.
Employees in the participating organisations received an e-mail inviting them to participate in the study
and provided them with a login name and password. Without these details, access to the survey could
not be obtained.

Questionnaire design

The questionnaires were designed to provide a greater understanding of the work-life balance
experiences of construction industry employees, and their domestic partners, in the private and public
sectors. Three main variables that have been widely investigated in the work-life area are: work
conditions and organisational environment; burnout; and organisational commitment (Allen, 2001;
Allen & Meyer, 1990, 1996; Benbow & Jolley, 2002; Chang, 1999; Dunham, Grube & Castaneda, 1994;
Herscovitch & Meyer, 2002; Jaros, 1997; Meyer, Stanley, Herscovitch, & Topolnytsky 2002; Selder &
Paustian, 1989; Sullivan, 1989; Yoon & Thye, 2002). However, it is very difficult to determine the
relationships between these factors without considering the impact of variables such as family and
relationship strain and support. In particular, the questionnaires investigated job and family
characteristics, work-life conflict and employee well-being, family functioning and work-related
attitudes. They were designed to measure the extent to which employees experienced tension between
work and non-work-life (work-life conflict), as well as antecedent and outcome variables found to be
associated with work-life strain in other industrial contexts.

Antecedents, or determinants of work-life conflict, included organisational or job characteristics, such

as job demand, work hours, work load and individual characteristics, such as parental responsibilities,
number of children etc. Outcome variables also included those experienced by the individual, such as
partner and parental relationships, family management and burnout, as well as variables known to have
a direct impact on organisational effectiveness. These included organisational commitment, turnover
intent and job satisfaction. It is important to recognise that individual outcomes, including burnout, are
closely linked to organisational outcomes. Several variables were included in our study as moderator
variables. These included perceived organisational support, supervisor and co-worker support and
flexible work arrangements.

The variables measured and the scales used for both the employee and the partner questionnaires are
presented below.

22
Employee Questionnaire

The construction industry employees questionnaires consisted of eight main sections, which
investigated the following areas:

Demographics;

Work load and responsibility;

Work environment;

Feelings about work;

Quality of spouse/partner relationship;

Family dependants;

Absence from work; and

Work-life balance.

In order to ensure a match between their own and their partners information, as well as allowing for a
potential longitudinal study in the future, respondents were asked to generate a 5 character code by
using the first 5 letters of their mothers maiden name. In addition, they were asked to indicate their
partners postcode as at 30 June 2003. This was to ensure anonymity.

Part A Demographic information

Demographic information collected from respondents included:

Year of birth;

Age;

Gender;

Number of years worked in the construction industry;

Number of years worked for their current employer;

Job position;

Location of the majority of their work time; and

Description of their family and household including dependent children

Household classifications included separate categories for partnered and non-partnered respondents
with or without dependent children. Dependent children were classified as those under the age of 18
years. These classifications were used to describe all forms of family structures, including non-
traditional forms, such as single-parents and couples with responsibility for children from previous
relationships (Rothausen, 1999).

Respondents were also asked to indicate their country of birth and the language they predominantly
spoke at home. However, in order to allay respondents concerns about providing detailed demographic
information which could identify individual employees, the provision of information about ethnicity
was made optional.

Part B Work load and responsibility

Work load
Major, Klein and Ehrhart (2002) suggest that overload occurs when the magnitude of work overwhelms
an individuals perceived ability to cope. A person may experience work overload even if the work is
completed on time. In this regard, a subjective evaluation of overload is more concerned with
appraising ones perceived ability to meet demands, and therefore captures the feeling state.

Subjective quantitative work load was assessed in this study via nine items scored on a 5-point Likert
style scale (Caplan, as cited in Cook, Hepworth, Wall & Warr, 1981). Participants were asked to
indicate the extent to which statements reflected the time demands of their work, where 1 = very little;
2 = little; 3 = some; 4 = great; and 5 = very great. The nine items included statements such as the
number of projects and/or assignments you have; and the extent to which you feel you never have
any time. Two items were reversed scored (items 3 and 6).

23
Sense of responsibility
For many people in the workplace, a source of strain is the high level of perceived responsibility.
Previous research with anaesthetists, surgeons, and air-traffic controllers demonstrated that a
combination of high attentional demand and responsibility in the workplace led to adverse
physiological and psychological outcomes (Cob & Rose; Gopher & Donchin; Payne & Rick; Moray;
ODonell & Eggemeir, as cited in Martin & Wall, 1989).

In this study, responsibility for things and persons were assessed via two scales (Caplan, as cited in
Cook, Hepworth, Wall & Warr, 1981). The first, responsibility for things included four items (items
1, 2, 3 and 4 in our scale). The second responsibility for persons consisted of two items (items 6 and
7). An additional item, the responsibility you have for the safety of others, was included, as the
authors recognised the importance of this factor for construction industry employees. Participants were
asked to consider the statements and to indicate the extent to which they accurately reflected their level
of responsibility at work. Examples of items included the responsibility you have for the work of
others; and the responsibility you have for budgets and expenditures. All three scales were scored
on a 5-point Likert scale (1 = very little; 2 = little; 3 = some; 4 = great; and 5 = very great). Caplan
based these scales of responsibilities on the work of Kahn, Wolfe, Quinn and Snoek (as cited in Cook,
Hepworth, Wall & Warr, 1981).

Work hours
Two measures relating to work hours were used in this study. Participants were first asked to indicate
the actual number of hours spent directly undertaking work duties. However, this did not take into
account the number of hours spent travelling to work (particularly if the worksite location varied); nor
did it consider time spent contemplating work. For this reason, participants in this study were asked to
indicate the average number of hours per week worked in the organisation and the number of hours
spent away from home due to work on a daily basis.

While self-reports of work hours are not always accurate, the perception of overload rather than actual
overload is of more interest, as it is when perceived work exceeds perceived ability to cope, that
stress occurs.

Part C Your work environment

Participants considered the most recent construction project on which they had worked and reported
the major features of it. In particular, questions were asked about type of project (building construction,
civil engineering construction or other); total construction cost; degree of technical complexity (low,
moderate and high); and whether or not they would choose to work on a similar project in the future.

Organisational support
In this study, perceived organisational support was measured via 16 items drawn from a larger 36-item
scale (Eisenberger, Huntington, Hutchison & Sowa, 1986). Items were scored on a 7-point scale where
1 = strongly disagree; 2 = disagree; 3 = slightly disagree; 4 = neither agree nor disagree; 5 = slightly
agree; 6 = agree; and 7 = strongly agree. Participants were instructed to read the statements about
organisations and decide the extent to which they agreed with them in relation to their own
organisation. It included items such as the organisation values my contribution to its well-being; and
the organisation disregards my best interests when it makes decisions that affect me (reverse scored).
Seven items were reverse scored (items 2, 3, 5, 6, 9, 12 and 13).

Supervisory support
In this study, supervisory support was measured utilising 13 items drawn from two larger scales
(Lambert, 2000; Thomas & Ganster, 1995). The first seven items were taken from an 8-item sub-scale
measuring supervisor support, originally drawn from the Michigan Assessment of Organizations
Questionnaire (Cammann, Fichman, Jenkins & Klesh & Families and Work Institute, as cited in

24
Lambert, 2000). The eighth item, my supervisor is concerned about the way we workers think and
feel about things, was excluded from the modified scale as it felt by the researchers that this issue was
adequately investigated in the additional items drawn from Shinn (as cited in Thomas & Ganster, 1995).
The term immediate supervisor was defined as the person who most often officially assessed their
job performance.

The scale included statements such as my immediate supervisor is understanding when I have personal
or family problems which interfere with my work, and my immediate supervisor appears to know a
lot about company policies that help employees manage their family responsibilities. The final six
items were taken from a nine-item supervisor support scale by Shinn (as cited in Thomas & Ganster,
1995). In the original scale, a 5-point frequency scoring method was used whereby participants were
asked to rate how often in the past two months their supervisor engaged in specific supportive
behaviours. However, in the modified supervisory support scale, participants were asked to read a list
of behavioural attributes and then indicate the frequency with which their immediate supervisor
behaved in that way. The statements included items such as my immediate supervisor switched
schedules (hours, overtime hours, vacation) to accommodate my family responsibilities, and was
scored as follows: 1 = not at all; 2 = rarely; 3 = some of the time; 4 = most of the time; and 5 = all of
the time. Three items were reverse scored (items 8, 11 and 13).

Co-worker support
Two scales were used to tap into practical and emotional support employees felt they received from
their co-workers. Co-workers are defined as the people with whom they had the most contact with in
the company, excluding their immediate supervisor. Faith in peers (Cook & Wall, as cited in Cook,
Hepworth, Wall & Warr, 1981) was measured via three items. Participants were asked to read the
statements about their work group, and to indicate the extent to which they agreed or disagreed with
them in relation to their co-workers. Responses were scored on a 7-point scale where 1 = no, I
strongly disagree; 2 = no I disagree quite a lot; 3 = no, I disagree just a little, 4 = Im not sure; 5 = yes, I
agree just a little; 6 = yes, I agree quite a lot; and 7 = yes, I strongly agree. All items were scored in the
same direction and summed. A high score was indicative of the perception of a high level of practical
support from co-workers.

Taylor and Bowers (as cited in Cook, Hepworth, Wall & Warr, 1981) developed a Survey of
Organisations Questionnaire, which included a component examining aspects of leadership. As part of
this, they included a number of peer leadership sub-scales, including a three-item peer support scale.
Participants were asked to read the statements and indicate the extent to which the statement reflected
the behaviour of their co-workers. Items were scored in a positive direction on a 5-point Likert scale,
where 1 = to a very little extent; 2 = to a little extent; 3 = to some extent; 4 = to a great extent; 5 = to a
very great extent. It included items such as how friendly or easy to approach are the persons in your
work group?. A high score was indicative of the perception of a high level of emotional support from
co-workers.

Part D Your feelings about work

Organisational commitment
Eighteen items evaluating commitment to the organisation were drawn from Meyer, Allen and Smiths
(1993) six factor questionnaire. In the original survey, the first three factors represented commitment to
the occupation, and the remaining three factors, commitment to the organisation. The first factor,
affective commitment to the organisation consisted of items numbered: 1, 4, 6, 7, 12, and 14 (items 4,
6 and 14 were reverse coded). Continuance commitment consisted of items numbered 3, 8, 10, 11, 13,
and 18, with item number 18 being reverse coded. The final factor (items 2, 5, 9, 15, 16, and 17),
normative commitment, contained only one reverse coded item (item 2).

Participants were asked to read the statements of work-related feelings and decide the extent to which
they agreed or disagreed with them in relation to their organisation. The scale contained statements
such as I would feel guilty if I left my organisation now, and I do not feel any obligation to remain

25
with my current employer. They were scored on a 7-point scale, where 1 = strongly disagree; 2 =
moderately disagree;, 3 = slightly disagree; 4 = neither agree nor disagree; 5 = slightly disagree; 6 =
moderately agree; and 7 = strongly agree. Mean scores for each of the three commitment scales were
obtained. A high score indicated a high degree of commitment to the organisation.

Turnover intentions
Turnover intentions has been defined as a general tendency to remain with or leave the
organisation (Whitener & Walz, as cited in Jaros, 1997). In the current study, it was measured via the
two items I often think about quitting and I will probably look for a new job in the next year,
which were drawn from a 3-item sub-scale of the Michigan Organizational Assessment Questionnaire
(Cammann, Fichman, Jenkins & Klesh; Sheashore, Lawler, Mirvis & Cammann, as cited in Cook,
Hepworth, Wall & Warr, 1981). They were scored on a modified 7-point scale where 1 = strongly
disagree; 2 = disagree; 3 = slightly disagree; 4 = neither agree nor disagree; 5 = slightly agree; 6 = agree,
and 7 = strongly agree. A higher mean score reflected a higher likelihood of a person leaving their job.

Job satisfaction
Job satisfaction was measured using a scale developed at the University of Michigan (Taylor & Bowers,
1972), and consisted of three sub-scales: intrinsic rewards (items 1, 2, and 3); extrinsic rewards (items 4,
5, and 6); and social rewards satisfaction (items 7, 8, and 9) were drawn from the Michiagan
Organizational Assessment Questionnaire (Cammann, Fichman, Jenkins & Klesh; Seashore, Lawler,
Mirvis & Cammann, as cited in Cook, Hepworth, Wall & Warr, 1981). The sub-scales included items
such as how satisfied are you with the chances you have to do something that makes you feel good
about yourself as a person?; how satisfied are you with the amount of pay you get?; and how
satisfied are you with the friendliness of the people you work with?. They were scored on a 7-point
scale, where 1 = very dissatisfied; 2 = dissatisfied; 3 = slightly dissatisfied;, 4 = neither satisfied nor
dissatisfied; 5 = slightly satisfied; 6 = satisfied, and 7 = very satisfied. For each of the three sub-scales,
the mean value represents the satisfaction score.

Burnout
This study utilised the 16-item Maslach Burnout Inventory (MBI) which comprises three sub-scales
assessing: emotional exhaustion (I feel emotionally drained from my work); cynicism (I have become
less interested in my work since I started this job); and professional efficacy (At work, I feel confident
that I am effective at getting things done). The items for the third dimension of burnout are framed in
positive terms and thus a low score reflects a low sense of professional efficacy. The MBI was selected
for use in the current study due to its brevity and proven reliability and validity of results (Corcoran,
1995; Shinn, 1982; Gaines and Jermier, 1983).

Since the response formats of intensity and frequency have been found to be highly correlated, only
frequency ratings were used. Items were rated on a 7- point Likert scale where 0 = never; 1 = a few
times a year or less; 2 = once a month or less; 3 = a few times a month; 4 = once a week; 5 = a few
times a week; and 6 = every day.

Part E Spouse/partner relationship

Work-life experiences of people in dual-income couples is likely to be different to those in single
income households. Participants were asked if they currently had a spouse/partner, whether their
partner worked in paid employment and, if so, how many hours they worked. This was assessed via a
categorical scale, with five possible responses as follows: 1 - 14 hours; 15 - 29 hours; 30 - 34 hours;
35 - 49 hours; and 50 or more hours.

Companionship, sociability and relationship tension

Participants were asked a number of questions relating to happiness in their relationship with their
spouse/partner. Orden and Bradburn (1968) constructed a list of events that occur in a marriage that

26
could be considered pleasurable or likely to cause disagreements. They contended that relationship
happiness occurs as a product of the balance between satisfactions and tensions.

Both participants and their partners were administered the Orden and Bradburn scales. The items in the
satisfactions list clustered into two distinct groups which original authors labelled marriage sociability
(renamed relationship sociability in this study) and marriage companionship (renamed
companionship with partner). Participants were provided with a list of activities that were considered to
be pleasurable and asked to rate how frequently they had undertaken them. A sample item from the
relationship sociability scale was visited friends together, and from the companionship with partner
scale had a good laugh together or shared a joke.

The relationship tension scale contained items about which people sometimes disagreed. Participants
were asked to indicate how frequently they had had a difference of opinion over these. A sample item
from this scale was household expenses. In this study, both the pleasurable and tension-provoking
items scored in the same direction, and the frequency of occurrence rating used was the same, 0 = not at
all; 1 = very infrequently; 2 = infrequently; 3 = frequently; and 4 = very frequently.

Social undermining
Perceived social undermining was assessed via 15 items which were drawn from Abbey, Abramis and
Caplan, and Westman, Etzion and Danon (as cited in Westman, Etzion & Danon, 2001, via a personal
communication). Participants were asked to read the questions and to indicate the frequency with
which they occurred, where: 1 = never; 2 = sometimes; and 3 = always (items 2, 4, 6, 8, 10, 12, and 14
were reverse scored).

Part F Family dependants

Family dependants
A broader view of family was utilised in this study so that the responsibility for children as well as other
family members (aged or disabled) could be assessed. Rothausens (1999) Responsibility for
Dependants scale (RFD) was used. This scale apportioned weightings for level of responsibility to
children and other dependants and factored in age, living arrangements and degree of supervision
required where dependants were disabled or dependent adults (i.e. aged parents). (See Table 1 for
details on weightings attached to each person within the family).

The RFD scores were not calculated, or used, within this report, but will be used in future publications
emanating from this study. Parental status and number and age of children were considered important
by the study task force and more relevant to this report.

Table 1 Responsibility for dependants (RFD) measure item and item weights

Item Weights
Final Items Living with Not living with
individual individual
Each child aged < 1 year 7.0 4.0
Each child aged 1 2 years 6.5 3.5
Each child aged 3 5 years 6.0 3.5
Each child aged 6 9 years 5.0 3.0
Each child aged 10 14 years 5.0 3.0
Each child aged 15 18 years 4.5 2.5
Each child aged over 18 (or) any other adult 3.0 1.0

For dependent adults or children with disabilities, add:

Supervision required 65 100% of time 7.0 3.5
Supervision required 25 64% of time 6.5 3.0
Some help needed but largely self-sufficient 5.0 2.5

27
Parenting behaviours
Robinson et al. (1995) created a 133-item scale which was considered representative of the three main
parenting typologies, authoritative; authoritarian; and permissive parenting practices. Due to space
restrictions, selected items from the authoritative and authoritarian sub-scales were selected. In
particular, items from the warm and involvement and good natured/easy going sub-scale of the
authoritative construct and verbal hostility from the authoritarian construct were selected. It was
considered these affection and hostility issues, in particular, could be subject to interference from work
stress and time commitments of parents.

Family support
Family support either from family and friends was determined by presenting the participants with four
statements regarding the level of support they believed they were receiving. This scale was developed
by the authors and was used in the previous report for the CIIA (Lingard & Francis, 2001). Possible
responses ranged from 1 (I/we have no family or friends who could look after the children) to 4 (I/we
can always get family or friends to look after the children). Participants were asked to tick the response
that best described the level of support they could rely on, with a higher score indicative of a higher
level of support.

Part G Absence from work

Absenteeism
Participants were asked to indicate the number of working days they were absent in the preceding
twelve months, excluding those taken for holiday purposes.

To explore reasons for this absenteeism, participants were asked to indicate how many of these days
were taken due to their own illness, to look after another who was unwell, and other reasons.
Respondents were asked to indicate what arrangements were made for this time off work, for example,
did they take sick leave, use annual leave or arrange to make up the time at a later date?

Annual leave
Participants were asked how many annual leave days they had off as holidays during the past 12
months. Research suggests that respite from a demanding job can alleviate some of its negative
outcomes, such as burnout (Westman & Eden, 1997). However, when faced with a tight project
schedule, it may be difficult for construction industry employees to take annual leave entitlements
regularly.

Part H Work-life balance

In this section of the questionnaire respondents were asked about the extent of their work-life conflict,
family management issues, control over work and family matters and their preferences for work-life
balance initiatives.

Work-life conflict
In this study, the scales for work-to-family and family-to-work conflict were drawn from the work of
Boles et al., which was based on the scales of Netemeyer, Boles and McMurrian (1996). Participants
were asked to read 10 statements about their experiences at work, and outside of work and to indicate
the extent to which they agreed or disagreed with them. The items were scored on a 7-point scale where
1 = strongly disagree; 2 = disagree; 3 = slightly disagree; 4 = neither agree nor disagree; 5 = slightly
agree; 6 = agree; and 7 = strongly agree. It included statements such as the demands of my work
interfere with my home and family life and my co-workers and peers at work dislike how often I am
preoccupied with my family life.

28
Family management
Family management was measured via a 21-item scale, designed by Bohen and Viveros-Long (1981).
Participants were asked to read a list of activities and to indicate the extent to which they believed it was
easy or difficult to perform them in their present job. The first eight items were for parents, for
example, to stay home at home with a sick child, whereas the next 13 were relevant for all
participants to go on errands (i.e. Post office, car service). The child-related items included a not
applicable response. The order of the scale was rearranged by the authors to increase ease of response,
and all items were score in the same direction.

Work family control

In this study, work-family control was measured via a 16-item scale. The first 14 items were developed
by Thomas and Ganster (1995). The additional two items were adapted from the work of Thomas and
Ganster, as the authors considered that it was important to evaluate level of responsibility in areas of
life other than in relation to dependants.

Participants were asked to read the list of questions about choices made in working life and to indicate
the degree of choice they felt they had in those areas, where 1 = very little; 2 = little; 3 = some; 4 =
great, and 5 = very great. In the original scale, the corresponding responses for 4 and 5 were much
and very much respectively. The first six items had an additional possible response; 8 = not
applicable. The questions included items such as how much choice do you have over the amount and
quality of care available for a sick child? and how much choice do you have over when you take
vacations or days off?

Work-life balance initiatives

In this study, the utility of work-life benefits were assessed via a 21-item composite scale. Participants
were asked to indicate the extent to which the work-life benefits would be of use, ignoring the
likelihood of future usage. Responses were scored on a 5-point Likert-style scale, where 1 = not at all
useful; 2 = of limited use; 3 = moderately useful; 4 = useful; and 5 = very useful. The items were
drawn from a wide variety of sources (Biggs, 1998; Fallon & Mallamace, 2000; Grandey, 2001; Gray &
Tudball, 2002; Hill, Hawkins, Ferris & Weitzman, 2001; Moen, Harris-Abbott, Lee & Roehling, 1999;
Saltzstein, Ting & Saltzstein, 2001; Scheibl & Dex, 1998; Thornwaite, 2002).

Additional comments
At the end of the questionnaire respondents were asked to indicate the five things they looked for in a
good employer. In addition, they were asked to indicate the five issues that would be taken into
consideration when deciding whether to leave their current employment to go to another employer.
Participants were asked whether they would be happy to be interviewed at some later stage and invited
to leave their contact details.

29
Spouse/partners questionnaire

The partners questionnaire contained six sections including:

Demographics;

Information relevant to employment;

Work-life balance;

Spouse/partner relationships;

Family dependants; and

Household management.

In addition, they were asked to indicate any other areas in which their partners employer could help
their partner to achieve a better balance between work and life.

In order to ensure a match between partner and employee information, as well as allowing for a
potential longitudinal study in the future, respondents were asked to generate a five-character code by
using the first five letters of their partners mothers maiden name. In addition, they were asked to
indicate their partners postcode as at 30 June 2003. This was to ensure anonymity of the respondent.

Part A Demographic information

The demographic information requested consisted of questions. The section started by asking the
participants year of birth and gender. Some employment questions followed relating to whether or not
the participant was in paid employment and, if so, their average number of work hours per week. Their
current job classification was ascertained using standard Australian Bureau of Statistics categories such
as: manager/administrator; professional; associate professional; tradesperson etc.

Part B Your employment

Absenteeism and annual leave

Participants who worked in paid employment were asked to estimate the number of days they took as
holidays during the previous 12 months; how many times a year they took off seven or more
consecutive days; and the number of days taken off in the previous 12 months to look after a
dependant.

Work involvement
Work involvement was measured via a 12-item scale where respondents were asked the extent to which
they agreed or disagreed with statements regarding how important work was to them. It included items
such as my chosen line of work gives me a sense of well-being and I have very strong ties to my
job. The first three items and were reverse coded. The remaining 9 items were adapted from Kanungo
(items 4 - 12) (1982) and measured work involvement. In the original questionnaire there were six
possible responses: 1 = strongly disagree; 2 = disagree; 3 = mildly disagree; 4 = mildly agree; 5 = agree;
and 6 = strongly agree. In the current study (where 1 = strongly agree; 2 = agree; 3 = neither agree nor
disagree; 4 = disagree; and 5 = strongly disagree), all items (except item 5) were reverse coded, and
summed. Thus, a higher score reflected an individuals view that their work was very important to
them.

Burnout
Once again, Maslachs Burnout Inventory (Maslach, Jackson & Leiter, 1996) was administered (see
Employee Questionnaire). The 16-item inventory comprises three sub-scales assessing: emotional
exhaustion (I feel emotionally drained from my work); cynicism (I have become less interested in my
work since I started this job); and professional efficacy (At work, I feel confident that I am effective at
getting things done).

30
Part C Work-life balance

Work-life balance control

The degree of control over balancing work with dependents care other interests, and other
responsibilities was ascertained via a 3-item scale (adapted from Bohen & Viveros-Long, 1981; and
Thomas & Ganster, 1995). Each of the items could be scored on a 5-point Likert-style scale where 1 =
very little; 2 = little; 3 = some; 4 = great; and 5 = very great. The first item, in general, how much
control do you have over the way you balance working and dependant care responsibilities?, also had
the option of not applicable. Participants were asked to read the list of questions about choices made
in working life and to indicate the degree of choice they had in those areas.

Part D Spouse/partner relationship

Companionship, sociability and relationship tension

Relationship happiness was assessed utilising the Orden and Bradburn (1968) list of pleasurable and
conflict provoking events. These scales tapped into three different constructs namely: companionship
with partner; relationship sociability; and relationship tension. They were also administered to industry
employee participants. (See employee questionnaire section above).

Social undermining
Perceived social undermining was again assessed via 15 items which were drawn from Abbey, Abramis
and Caplan, and Westman, Etzion and Danon, (as cited in Westman, Etzion & Danon, 2001, via a
personal communication). (See Employee Questionnaire).

Relationship affected by partners work commitments

Participants were asked to indicate the extent to which their relationship was affected by their partners
work commitments. A 5-point Likert scale was used with scores ranging from very little (1) to very
great (5).

Part E Family dependants

Participants were asked whether they had any dependant children. Dependent children were defined as
being under 18 years of age.

Parenting behaviours
The 16 items from the Robinson et al. (1995) scale relating to specific parenting behaviours were used.
These were considered to be subject to interference from work stress and time commitments of
parents. (See Employee Questionnaire).

Part F Household management

A modified version of Baxters (1993) measure, The Gender Division of Domestic Labour, was used
in this section. The items included in this study related to both routine household work (both indoor
and outdoor housework), as well as tasks relating to the care of children. These are discussed below. In
this study, Baxters measure has been used in order to look beyond the time commitment per week.

Time commitment to parental activities

Participants were asked about their and their partners, time commitment to taking care of, or doing,
things with their children (i.e. feeding, dressing, washing, going to places, helping with homework,
disciplining, talking, reading, driving them to places etc). The items provided a breakdown of the
number of hours both parents were involved in childcare on both working and non-working days.

31
Typical household tasks
The scale employed in the study had three items relating to indoor housework (preparation of meals,
cleaning of house, and grocery shopping) and three items relating to outdoor housework (puts out
garbage, mows the lawn, and general maintenance/ improvement). Three aspects of each task were
considered, namely: who usually performs the specific household tasks; the frequency with which they
were performed; and how much time is spent on each of them.

Responses relating to who performs the housework task where 1= self always performs task; 2 = self
usually; 3 = self and partner equally; 4 = partner usually; 5 = partner always; 6= other please specify; 7
= not applicable. Responses relating to how often you perform this task where 1 = several times a day;
2 = once a day; 3 = several times a week; 4 = once a week; and 5 = less than once a week. Participants
were also asked to indicate their time commitment to the particular task in an average week.

Additional information
Participants were asked to indicate any areas in which their partners employer could help them achieve
a better work-life balance. Participants were asked whether they would be happy to be interviewed at
some later stage and invited to leave their contact details.

Data analysis

In the first instance, the reliability of our measurement of the variables was examined and descriptive
statistics for each variable, such as means and standard deviations, were calculated.

Reliability of measurement
Reliability analysis allows an assessment of the measurement scales and items that make them up. For
example, reliability analysis provides an answer to the question does my questionnaire measure work-
life conflict in a useful way?. Using reliability analysis, the extent to which items in the questionnaire
scales are related to each other is considered and an overall index of the internal consistency of each
scale is produced. Problem items that should be excluded from the scale can also be identified.
Cronbachs alpha coefficients were calculated for each of the scales. Cronbachs alpha is a measure of
internal consistency based on the average inter-item correlation. These reliability and descriptive
statistics are presented in the results. The descriptive statistics enable an assessment of the overall levels
of work-life conflict experienced by respondents and the antecedents and outcomes of this conflict to
be assessed.

However, these descriptive statistics do not allow enable the experiences of different demographic
groups, such as respondents working in the public and private sectors, to be compared. Neither do
these statistics provide an insight into the way in which the variables are linked. For example, it would
be useful to know whether, and in what way, work schedule demands are associated with work-life
conflict. It is also helpful to find out whether, and in what way, work-life conflict is associated with
individual, family or organisational outcomes, such as burnout, relationship quality or job satisfaction.
In order to explore some of these differences and linkages between variables, further statistical analysis
was undertaken using the statistical comparison of means, correlation analysis, multiple regression and
modelling techniques described below.

Factor analysis
Some variables measured in the study are known to be multi-dimensional. For example, previous
research has demonstrated that organisational commitment is not a one-dimensional concept but is
made up of three distinct factors. Similarly, research suggests that the quality of relationships with
other family members is multi-dimensional. For this reason, principal components factor analysis was
used to analyse the questions included in the questionnaire to measure multi-dimensional variables,

32
including supervisor support, organisational commitment, burnout, family management and work-life
benefits.

Factor analysis assumes that underlying dimensions or factors can be used to explain complex
phenomena. The goal of factor analysis is to identify the not-directly-observable factors based on a set
of observable or measurable indicators (Norusis 1993). Norusis describes the process of factor analysis
as follows. The first step in factor analysis is to produce a correlation matrix for all variables. Variables
that do not appear to be related to other variables can be identified from this matrix. The number of
factors necessary to represent the data and the method for calculating them must then be determined.
Principal components analysis is one method of extracting factors. In principal components analysis,
linear combinations of variables are formed. The first principal component is that which accounts for
the largest amount of variance in the sample, the second principal component is that which accounts
for the next largest amount of variance and is uncorrelated with the first and so on. At this step it is
also necessary to ascertain how well the model fits the data. Coefficients (factor loadings), that relate
variables to the identified factors were calculated. In order for an item to belong to a given factor it is
recommended that the loading value be no less than 0.40.

The factor model is then rotated to transform the factors and make them more interpretable. The
rotation phase transforms a factor matrix in which most factors are correlated with many variables into
one in which each factor has non-zero loadings for only some of the variables. The most commonly
used method for rotation is varimax rotation which seeks to minimise the number of variables that have
high loadings on a factor thus permitting the factors to be differentiated from one another. Following
rotation, scores for each factor can be computed for each case in a sample. These scores can then be
used in further data analysis, such as analyses of variance, correlation and regression analysis.

The results of the factor analysis are presented in the results, alongside the reliability information for
the resultant factors.

Statistical comparisons of means

In order to compare the work-life experiences of different groups of employees, statistical tests were
conducted to compare the mean scores for important variables between groups of employees. Analyses
of variance (often abbreviated to ANOVAs) and independent t-tests were used to test for significant
differences.

Correlation analysis
In order to determine the nature and strength of linkages between the variables measured, bi-variate
correlation analyses were conducted. These correlations enable an assessment of the degree to which
one variable is linearly related to another. Correlation coefficients were calculated for each pair of
variables measured in the study. These statistics are presented under the sub-heading Bi-variate
Correlations in section 3 of this report. The correlation coefficients, presented in matrix format,
indicate the strength and direction of the linear association between two variables. A positive
coefficient indicates that as one variable increases, so too does the other variable. A negative coefficient
indicates an inverse relationship, i.e. as one variable increases, the other decreases. Two-tailed
significance levels associated with the correlation coefficients were also calculated. Thus, if r=.088 and
this is significant at the 0.01 level, the probability of obtaining a correlation coefficient of at least 0.88 in
absolute value, when there is no linear association between the two variables, is less than 1 %.
Correlation analyses were undertaken as a precursor to the multiple regression analyses because, owing
to the large number variables measured, only those which were significantly associated with the
important outcome variables were included in the regression models.

Multiple regression analysis

Having gained an understanding of the associations between the variables, it is useful to determine to
what extent one variable can be predicted, given knowledge of another variable. Modelling procedures
together with multiple regression analyses were undertaken to determine the extent to which selected

33
antecedent variables predicted certain outcome variables. Several models were developed, based on
existing theory of work-life issues. Once models are specified, data are analysed to determine whether
the model is consistent with them. Failure of the model to fit the data results in model falsification,
whereas a good fit supports the theoretical arguments. Although no model can be definitively
confirmed, the repeated failure to disprove a model adds strength to researchers belief in the theory
(Cohen et al., 2003).

The theory-driven models we developed relating to work-life interface were tested to determine the
extent to which, in our sample, specified outcome variables were explained by the other variables
measured in our study. These models and the statistical results are presented in the later part of the
employee questionnaire results in section 3 of this report. Multiple regression analysis also enables the
relative importance of independent variables to be determined. For example, multiple regression can be
used to answer the question: how important are work schedule flexibility and work schedule demands
when they are used to predict work-life conflict along with other independent variables in the
regression equation?. Beta coefficients (denoted ) for the independent variables in a regression model
can be compared. Beta coefficients are standardised to accommodate differences between units of
measurement among the independent variables. Another way of assessing the relative importance of
independent variables is to consider the increase in R2 when a variable is entered into an equation that
already contains the other independent variables. R2 indicates the percentage of variation in the
dependent variable that can be explained by the regression model. Thus, an R2 of 0.77 indicates that
77% of variation in the dependent variable is explained by the model.

Moderators and mediators

There are a number of different roles variables can play in theory. A cause (X) of some variable (Y) is
believed to precede Y in time. However, some theories provide for the inclusion of mechanisms
through which variable X is related to variable Y. For example, previous studies have shown that work
hours are related to burnout through work-to-family conflict (Lingard & Francis, in press). These
intervening variables are called mediators of the effect of X on Y. Figure 1 (a) provides an example of a
mediator (Z) that totally accounts for the relationship between X and Y. Figure 1 (b) shows a mediator
(W) that partially accounts for the relationship between X and Y.

Figure 1 Mediating variable Z and W (Source, Cohen et al., 2003, p. 458).

In order to test for mediation effects we used procedures described by Baron and Kenny (1986). Baron
and Kenny suggest that, to test for mediation, three regression equations must be estimated, as follows:
1. the mediator is regressed on the independent variable;
2. the dependent variable is regressed on the independent variable; and
3. the dependent variable is regressed on both the independent variable and the mediator.

34
To establish mediation, the independent variable must affect the mediator in the first equation; the
independent variable must affect the dependent variable in the second equation; and the mediator must
affect the dependent variable in the third equation. If these conditions hold, then the effect of the
independent variable on the dependent variable in the third equation must be less in the third equation
than in the second.

Other important variables in some causal theories are moderators. These are variables which modify the
relationships among other variables (see Figure 2). The arrow from Z to the arrow from X to Y
indicates that the estimate of the causal effect of X on Y is conditional on the value of Z.

Figure 2 Contingent effects, Z as a moderator (Source, Cohen et al., 2003, p. 458).

In order to test for moderation effects, we used procedures described by Baron and Kenny (1986).
Prior to conducting any tests for moderation effects, all continuously measured predictor variables were
centred (Aiken & West, 1991). Centering is a linear transformation method which eliminates problems
associated with multi-collinearity. It is achieved by subtracting the mean value for a variable from each
score for that variable. The power of moderated regressions to detect interactions is reported to be low,
resulting in a high Type II error rate (Aiken & West 1991). Butler et al., (2004) suggest one remedy for
this is to accept a higher Type I error rate. As a result of this, we accepted an alpha level of .10 in
testing the significance of moderation effects.

35
36
Section 3: Results
Introduction

The results section is presented in two main parts relating to the employee and spouse/partner
questionnaires.

Employee questionnaire

A total of 176 respondents answered the on-line survey and 29 respondents chose to download the
PDF version of the survey available on the website and returned a completed hard copy to the research
team. Of the 205 completed questionnaires, 202 surveys were considered as part of the analyses with
63 (31.2%) respondents from the private sector organisation and 139 (68.8%) respondents from the
public sector organisation. This represents a response rate of approximately 60% from the private
sector organisation and 28% from the public sector organisation.

The employee section of the results contains a number of subsections:

Descriptive analysis including details relating to demographics and each scale utlised in the
survey;

Comparison of public and private sector differences using a series of One-way Analyses of
Variance (ANOVAs);

Bi-variate correlations for major work and family related variables;

Analysis of Burnout including tests for mediation;

Multiple regression analyses to establish interactions between independent variables; and

Integrated model of the work-family interface.

Demographic information
Table 2 shows the demographic characteristics of the sample. These are then discussed in more detail
below.

37
Table 2 Demographic characteristics of sample
N % N %
Age Job description
20 29 years 39 19.3 Site/project engineer 34 18.8
30 39 years 72 35.6 Project/construction manager 41 20.3
40 49 years 41 20.3 Contract administration 15 7.4
50 59 years 43 21.3 Foreman/supervisor 7 3.5
60 years + 7 3.5 Support services 25 12.4
Engineering services 16 7.9
Gender Corporate management 16 7.9
Male 178 88.1 Other 6 3.0
Female 24 11.9
Description of household
Years worked in construction Couple with dependent children 85 42.1
1 9 years 74 36.6 Couple with non-dependent children 31 15.3
10 19 years 51 25.2 Single parent 6 3.0
20 29 years 40 19.8 Couple without children 32 15.8
30 years + 37 18.3 Single person 48 23.8

Hours worked per week Spouse/Partner

0 29 hours 4 2.0 Yes 158 78.2
30 39 hours 26 12.9 No 43 21.3
40 49 hours 92 45.5
50 59 hours 44 21.8 Dependent children
60 hours + 36 17.8 Yes 103
No 99
Work location
On site 9 4.5 Country of birth
Site office 66 32.7 Australia 166 82.2
Head or regional office 126 62.4 Other 36 17.8

Age, gender and nationality of the sample

The mean age of the sample was 39.81 years (SD=10.705) ranging from 22 to 67 years. The sample was
separated into several age categories. The composition within the sample is illustrated in the figure
below (see Figure 3).

Figure 3 Age cohorts of all respondents

3.47%
60 years + 19.31%
20 to 29 years
21.29%
50 to 59 years

20.3%
40 to 49 years 35.64%
30 to 39 years

38
The age distribution of public and private sector employees was found to be reasonably similar with the
average age of public sector employees being 40.19 years, (SD=10.81) and private sector employees
38.97 years, (SD=10.51). See Figure 4.

Figure 4 Public and private sector employees by age cohorts

50.0%
47.6%

40.0%

Company
Private sector
30.2%
Percent

30.0% Public sector

26.6%

21.6% 20.6%
20.0%

14.3%

10.0% 9.5%
7.9%

1.4%
0.0%
20 to 29 years 30 to 39 years 40 to 49 years 50 to 59 years 60 years or
more
Age cohorts

Of the 202 respondents 178 (88.1%) were male and 24 (11.9%) were female. While the female cohort
were represented in similar proportions in both the public and private sector (see Figure 5) the average
age of the female cohort was substantially younger than that of the male cohort.

Figure 5 Male and female respondents by company

100.0%
88.9% 87.8%
80.0% Company
Private sector
Percent

60.0% Public sector

40.0%

20.0%
11.1% 12.2%

0.0%
male female
Gender

The average age of male respondents was 40.98 years (SD=10.535), and female respondents 31.08 years
(SD=7.569). See Figure 6.

39
Figure 6 Male and female respondents by age cohorts

60.0% 58.3%

50.0%

Gender
40.0% male
36.0%
Percent

33.3% female
30.0%
23.6%
22.5%
20.0%
14.0%

10.0%
4.2% 4.2% 3.9%

0.0%
20 to 29 years 30 to 39 years 40 to 49 years 50 to 59 years 60 years or more
Age cohort

Thirty-six (17.8%) respondents reported not being born in Australia; 15 of these were born in the
United Kingdom and five in New Zealand. The remaining 16 respondents nominated 16 different
countries of origin. The other countries included Fiji Islands, South Africa, Netherlands, Solomon
Islands, Singapore, India, USA, Nepal, China, Malaysia, Hungary, Sweden and Indonesia.

Figure 7 Australian-born respondents by company

100.0%
90.6%

80.0%
Company
63.5% Private sector
Percent

60.0%
Public sector

40.0% 36.5%

20.0%

9.4%

0.0%
yes no
Born in Australia?

Only five respondents reported speaking a language other than English at home. These languages
included Spanish, Hindi, Pijin (English), Nepali and Cantonese.

Work related information

Construction industry experience

The number of years that respondents had worked in the construction industry varied from 1 to 50,
with an average of 16.59 years (SD=11.437). Figure 8 illustrates the similarities between the
respondents in the private and public sector organisations in terms of industry experience. The average
number of years of construction industry experience of those in the public sector organisation was
17.26 years, (SD=11.714), and private sector organisation 15.11 years (SD=11.714).

40
Figure 8 Construction experience by company

39.7%
40.0%

35.3%

30.0%
27.0%
Company
24.5% Private sector
Percent

22.2% 21.6% Public sector

20.0% 18.7%

11.1%
10.0%

0.0%
less than 9 years 10 to 19 years 20 to 29 years more than 30 years

Years of construction experience

Due to the lower average age of female respondents it was not surprising that their level of
construction experience was less than that of their male counterparts. The average construction
experience of male respondents was 17.84 years (SD=11.491), and female respondents was 7.29 years
(SD=4.974). This is represented in Figure 9 below.

Figure 9 Construction experience by gender

80.0% 79.2%

60.0%
Gender
male
Percent

female
40.0%

30.9%
26.4%
21.9% 20.8%
20.0% 16.7%

4.2%

0.0%
less than 9 years 10 to 19 years 20 to 29 years more than 30 years
Years of Construction experience

Tenure
Tenure was measured by asking the respondents how many years they had worked for their present
employer. The responses ranged from 1 to 38 years with an average of 10.87 years (SD=10.164).
However, significant differences were observed in the public and private sector employee responses.
The average number of years working for the public sector organisation was 13.18 years (SD=11.069)
and the private sector organisation was 5.78 years (SD=4.887). This is illustrated in Figure 10.

41
Figure 10 Construction experience and tenure by company

20
Years worked in
17.26
construction industry
Years worked for
15.11 current employer
15

Mean 13.18

10

5.78
5

0
Private sector Public sector
Company

Work hours
Respondents were asked the average number of hours they worked per week and on a daily basis, the
number of hours they were away from home in the past seven days. This second category was only
answered fully by 67% of the sample; however, some of these responses were questionable in terms of
their accuracy. For instance, a number of respondents answered zero for each of the days making the
data questionable.

For the purpose of this report only the average number of hours worked will be considered. The
second category did, however, reveal a number of employees who were away from home on work-
related activities for substantial periods of time, indicating that they were not living at home on a
permanent basis. This group may be worthy of closer analysis at a later date.

The average number of hours per week worked by respondents was 47.40 with a standard deviation of
10.524 (range, 8 to 80).

The sample was divided into several categories of average hours worked per week namely: 0 to 29 hrs;
30 to 39 hours; 40 to 49 hours; 50 to 59 hours; and 60 hours or more. Figure 11 illustrates the
breakdown for the whole sample.

Figure 11 Breakdown of average hours worked per week.

17.82% 12.87%

Hours worked
per week
0 to 29 hours
30 to 39 hours
40 to 49 hours
50 to 59 hours
60 hours or
21.78% more

45.54%

42
On further analysis it is apparent that the average work hours in the private sector organisation was
higher than that of the private sector organisation. The average number of hours worked by those in
the public sector was 43.26 hours (SD=7.766), and private sector 56.52 years (SD=10.088). This is
illustrated in the figure below.

Figure 12 Average hours worked per week by company

60.0% 59.0%

50.0% 49.2%

Company
40.0% Private sector
Percent

33.3% Public sector

30.0%

20.0% 18.7%
15.9% 16.5%

10.0%

3.6%
1.6% 2.2%
0.0%
0 to 29 hours 30 to 39 hours 40 to 49 hours 50 to 59 hours 60 hours or
more
Hours worked per week

A further breakdown was undertaken for gender. The average hours worked per week for male
respondents was 47.79 (SD=10.690), and female respondents 44.50 (SD=8.856). This is represented in
Figure 13.

Figure 13 Average hours worked per week by gender

60.0%
54.2%

50.0%
44.4%
Gender
40.0% male
Percent

female

30.0%
25.0%
21.3%
19.7%
20.0%

12.9% 12.5%

10.0%
4.2% 4.2%
1.7%
0.0%
0 to 29 hours 30 to 39 hours 40 to 49 hours 50 to 59 hours 60 hours or more
Hours worked per week

43
Position description

Respondents were asked to identify which classification most closely described their current position.
These categories included: Site/Project Engineer; Project/Construction Manager;
Clerical/Administrative; Contract Administration; Foreman/Supervisor/Site Superintendent;
Support Services (Safety, Human Resources, Quality, Environmental etc); Engineering Services
(Design, Electrical, Mechanical, etc.); Corporate Management and a final category None of the
above. Please specify. A total of 28 respondents indicated they were in this final category; however on
further inspection most, with the exception of six, were accommodated within the given classifications.
The three respondents who indicated they were in clerical/administrative roles were excluded from the
sample as they were not considered to fit the description of project-based technical staff. Table 3 details
the respondents position descriptions for the sample as a whole; for each company; and for male and
female.

Table 3 Position description by sample, company and gender

Company Gender
Position Description Total Private Public
male female
sector sector
Site/Project Engineer 34 19 15 26 8
Project/Construction Manager 41 13 28 39 2
Contract/Administration 15 5 10 12 3
Foreman/Supervisor/Site Superintendent 7 4 3 7 0
Support Services 25 8 17 17 8
Engineering Services 58 12 46 56 2
Corporate Management 16 1 15 15 1
Other 6 1 5 6 0
Total 202 63 139 178 24

Figure 14 illustrates the different types of positions the respondents held within the private and public
sector organization. This clearly reflects the different nature of the organisations and the work undertaken.

Figure 14 Position description by company

40.0%

33.1%
30.2%
30.0%

Company
Percent

20.6% Private sector

20.0% 19.0%
Public sector

12.2%
10.8% 10.8%
10.0% 7.9%
6.3%

2.2% 1.6%
0.0%
Si

Su

ot
Pr

Fo

En
on

or

he
te

oj

re

pp

gi

po
/P

ec

tra

r
ne
m

or
ro

ra
t/C

an
ct

er
tS
je

te
/A

/S

in
on
ct

er
dm

M
g
up
st

vic
En

an
Se
ru

er
ini

es
gi

ag
rv
ct

vis
st
ne

io

ic

em
r

or
a

es
n
er

tio

en
M

n
an

t
ag
er

Position Description

44
While the sample consist of only 24 women it is clear from Figure 15 that they tend to be in
site/project engineer or support services roles (i.e. safety, human resources, quality, environmental etc.).

Figure 15 Position description by gender

40.0%

33.3% 33.3%
31.5%
30.0%

Gender
Percent

21.9% Male
20.0% Female
14.6%
12.5%

10.0% 9.6%
8.3% 8.3% 8.4%
6.7%
3.9% 4.2%
3.4%

0.0%
Si

Su

ot
Pr

Fo

En
on

or

he
te

re

pp

g
je

p
/P

in
t

r
m
ra

or
or
ct

ee
ro

an
ct

at
/C

tS
j

rin
ec

/A

e
/S
on

er
dm

M
g
t

up
st

vi
En

an
Se
ru

ce
er
in
g

ag
r
is
ct

vi

s
in

vi
tr a
io

so

em
ce
ee

r
tio

s
r

en
M

n
an

t
ag
er

Position description

Work location
Employees were asked to indicate where they typically work: On site - in direct construction activity,
On site - but mainly in the site office; or Head or regional office. These are summarised for the
sample in Table 4.

Table 4 Work location

Number Percentage

On site - in direct construction activity 9 4.5

On site - but mainly in the site office 66 32.8
Head or regional office 126 62.7
Total 201 100.0

The location of work for employees within the private sector organisation and public sector
organisation varied quite substantially. Again this can be attributed to the nature of the work
undertaken within these companies. As this study investigated project-based personnel what is apparent
is that, in the majority of cases, pubic sector respondents work in the regional or head office, and
private sector respondents at the construction site of the project they worked on.

45
Figure 16 Work location by company

80.0% 77.7%

59.7%
60.0%
Percent

Company
40.0%
Private sector
29.0% Public sector
20.9%
20.0%
11.3%

1.4%
0.0%
On site-in direct On site- but mainly in the Head or regional office
construction activity site office
Work Location

It is apparent that while none of the women in the sample were involved in direct construction activity,
the total proportion working on site and in regional/head office was quite similar to that of their male
colleagues.

Figure 17 Work location by gender

70.0%
64.4% Gender
60.0% male
50.0% 50.0% female
50.0%
Percent

40.0%

30.5%
30.0%

20.0%

10.0%
5.1%

0.0%
On site-in direct On site- but mainly in Head or regional office
construction activity the site office
Work location

Construction project type, cost and technical complexity

Of the 200 respondents who answered the question regarding the type of construction project they had
most recently been involved in, 68.8% indicated civil engineering construction, 10.4% indicated
building construction and 19.8% indicated other. When analyzing this at the company level it was
apparent that a large portion of employees within the public sector organisation involved in civil
engineering construction. This reflects the nature of the work this organisation provides within the
public sector. The private sector organisation employees were involved in a mixture of project types.

46
Figure 18 Type of project by company.

79.0%
80.0%

60.0%

48.4% Company
Percent

Private sector
40.0%
Public sector
32.3%

19.4% 20.3%
20.0%

0.7%
0.0%
Civil Engineering Building Construction Other
Construction
Type of project

The survey asked respondents to indicate the total construction cost of their most recent construction
project. The majority of respondents indicated projects within the AUD$20m to AUD$100m category.
A breakdown according to organisation type and gender is provided in Table 5. This clearly indicates
the work undertaken by the private sector is of higher construction cost.

Table 5 Project construction cost by company and gender

Private sector Public sector

male female male female
% % % %
Over AUD$100m 25.5% 14.3% 7.5%
Between AUD $20 and AUD $100 million 49.1% 85.7% 12.5% 25.0%
Between AUD$5 AUDnd AUD$20 million 21.8% 16.7% 6.3%
Between AUD$1 and AUD$5 million 1.8% 32.5% 12.5%
Between AUD$500,000 and AUD$1 million 10.8% 12.5%
Between AUD$100,000 and AUD$500,000 9.2% 31.3%
Between AUD$40,000 and AUD$100,000 3.3%
Less than AUD$40,000 1.8% 7.5% 12.5%

Participants were asked to indicate the technical complexity of this recent construction project. Most
respondents perceived their projects to be of moderate technical complexity, with only 17.2%
perceiving them to be of low complexity and 22.2% perceiving them to be highly complex.

47
Figure 19 Project complexity

17.17%
22.22% Low complexity
High complexity

60.61%
Moderate complexity

The breakdown of technical complexity reveals a pattern for slightly greater perceived project
complexity for employees within the private sector organisation than employees of the public sector
organisation. This may well be related to the higher, on average, project construction cost. This is
illustrated in Figure 20 below.

Figure 20 Project technical complexity by company

70.0%

62.5%

60.0%
56.5%

50.0%
Company
Private sector
Public sector
Percent

40.0% 37.1%

30.0%

22.1%
20.0%
15.4%

10.0%
6.5%

0.0%
Low complexity Moderate complexity High complexity
Technical complexity of most recent project

The breakdown of technical complexity reveals a similar pattern for perceived project complexity for
male and female respondents. This is illustrated in Figure 21.

48
Figure 21 Project technical complexity by gender

70.0%

60.6% 60.9%
60.0%

50.0%
Gender
male
female
Percent

40.0%

30.0%

21.7% 22.9%
20.0% 17.4%
16.6%

10.0%

0.0%
Low complexity Moderate complexity High complexity
Technical complexity of most recent project

Respondents were asked their project preferences. The majority (75.5%) indicated they were happy to
work on a similar project again; 18.3% a larger/more complex project; and only 3.5% preferred a
smaller/less complex project. This is seen in Figure 22.

Figure 22 Future project preference by company

100.0%

80.6%
80.0% 76.3%

Company
Percent

60.0%
Private sector
Public sector
40.0%

20.7%
20.0%
14.5%

4.8% 3.0%
0.0%
yes no, I'd prefer larger/more no, I'd prefer smaller/less
complex project complex project

While the number of female respondents was low, it is interesting to note that a greater percentage of
women than men would prefer to work on a smaller/less complex project; an equal percentage of
women, (17.4%) would prefer a different project option. This is illustrated in Figure 23.

49
Figure 23 Future project preference by gender
79.3%
80.0%

65.2%

60.0%

Gender
Percent

male
40.0%
female

20.0% 19.0% 17.4% 17.4%

1.7%
0.0%
yes no, I'd prefer no, I'd prefer
larger/more complex smaller/less complex
project project

Family-related information

Family structure
Participants were asked to nominate their family structure from a list of options provided. The majority
(201) of the 202 respondents did so and the results for the whole sample and breakdown for company
and gender are provided in Table 6.

Table 6 Family structure by company and gender

Company Gender
Total
sample Private Public
Male Female
sector sector
Couple with dependent children 42.1% 46.0% 40.3% 46.6% 8.3%
Couple with non-dependent children 15.3% 14.3% 15.8% 15.2% 16.7%
Single parent 3.0% 4.8% 2.2% 3.4% 0%
Couple without children 15.8% 17.5% 15.1% 12.4% 41.7%
Single person 23.8% 17.5% 26.6% 22.5% 33.3%

The family structures for public and private sector employees were similar. However, when gender was
taken into account some significant differences were noted.

Figure 24 Family structure by gender

50.0% 46.6%
Gender
41.7%
40.0%
male
female
33.3%
Percent

30.0%

22.5%
20.0%
16.7%
15.2%
12.4%
10.0% 8.3%
3.4%

0.0%
Couple with Couple with Single parent Couple without Single person
dependent non-dependent children
children children

Family structure

50
The percentage of female respondents without children was much lower than that of the male
respondents. The difference in age of respondents may account for these results.

Participants were asked whether they had a spouse or partner. The results are shown in Table 7.
Table 7 Spouse/partner status

Gender Company
Total
Male Female Private sector Public sector

Non partnered 21.4% 20.3% 29.2% 14.5% 24.5%

Partnered 78.6% 79.7% 70.8% 85.5% 75.5%

Spouse/partner employment
Of the 158 respondents who indicated they had a spouse/partner, 114 (72.2%) indicated that their
partner worked. While a slightly higher proportion of spouses of respondents employed in the public
sector organisation worked (see Figure 25) significant differences were seen in the employment patterns
of spouse/partners of the female respondents.
Figure 25 Spouse/partner in paid work by company

80.0% Company
73.3%
69.8%
Private sector
Public sector
60.0%
Percent

40.0%

30.2%
26.7%

20.0%

0.0%
no yes
Spouse/partner in paid employment

All female respondents who had a spouse/partner indicated that their spouse worked (see Figure 26).

Figure 26 Spouse/partner in paid work by gender

100.0%
100.0%

80.0%
68.8%
Gender
Percent

60.0% male
female
40.0%
31.2%

20.0%

0.0%
no yes
Spouse/partner in paid employment

The number of hours worked by the spouse/partner is presented in the Figure 27. The categories
provided were 0-14 hours; 15 29 hours; 30 34 hours; 35 49 hours; and 50 or more hours.
The results indicate that nearly half the spouse/partners who work do so on a part-time basis.

51
Figure 27 Hours worked by spouse/partner per week

8.77%
50 or more hours 17.54%
1-14 hours

19.3%
15-29 hours

45.61%
35-49 hours

8.77%
30-34 hours

On further analysis it is apparent that the employment patterns of the spouse/partners of the female
respondents were different from that of their male counterparts. A much higher proportion worked
fulltime, with only two female respondents indicating their spouse/partner worked in a part-time
capacity.

Figure 28 Hours worked by spouse/partner per week by gender

50.0%
47.1%
45.4%
41.2%
40.0% Gender
male
female
Percent

30.0%

21.6%
19.6%
20.0%

10.3%
10.0%
5.9% 5.9%
3.1%

0.0%
1-14 hours 15-29 hours 30-34 hours 35-49 hours 50 or more hours
Partner/spouse hours worked per week

Number of children

Ninety-nine (of 199 respondents) indicated they had children. The average number of children was 2.19
(SD= 0.981).
Table 8 Number of children.
No of children Number %
1 child 27 26.2
2 children 42 40.8
3 children 22 21.4
4 children 11 10.7
5 children 1 1.0

52
Parents within the public sector organisation had on average more children (2.29 children, SD=1.023)
compared to their private sector counterparts (2.00 children, SD=0.874). See Figure 29.

Figure 29 Number of children by company

50.0%

42.9%
39.7%
40.0%
Company
31.4% Private sector
Percent

30.0% Public sector

23.5%
22.1%
20.0%
20.0%

13.2%

10.0%
5.7%

1.5%
0.0%
1 2 3 4 5
Number of children

Only three female participants reported having children. No women within the sample had school-aged
children.

Figure 30 Number of children by gender

70.0% 66.67%

60.0%
Gender
50.0% male
female
41.0%
Percent

40.0%
33.33%
30.0%
25.0%
22.0%
20.0%

11.0%
10.0%

1.0%
0.0%
1 2 3 4 5
Number of children

Absence from work

Respondents were asked to indicate the number of days they had taken off work in the past year as
annual leave, as well as days taken for other purposes.

The average number of days taken off work for non-holiday purposes ranged from 0 to 90 with a mean
score of 6.73 days (SD=12.429). Respondents were then asked the reason for this absence. One
hundred and fifty indicated they had taken leave due to sickness. The average number of days was 5.29
(SD=11.29). Eighty-five respondents indicated they took some leave to care for a sick dependant with
the average being 2.51 days (SD=5.500).

53
Respondents were asked to indicate how many days they had taken for holiday purposes. The average
for this purpose was 16.42 days (ranging from 0 to 121 days, SD=14.397). It is interesting to note that
while no gender differences were found the average for public and private sector employees was
somewhat different. This is illustrated in Figure 31.

Figure 31 Days off in past 12 months by company

20.00
17.85 No of days absent
(excluding annual
leave)
15.00
13.18
No of annual leave
days
Mean

10.00
8.32

5.00
3.21

0.00
Private sector Public sector
Company

A further break down of annual leave days indicates that 47.7% of private sector employees were taking
less than 10 days annual leave per year. The Figure 32 provides further information on the amount of
annual leave being taken by public and private sector employees.

Figure 32 Annual leave taken by company

50.0%
46.0%

42.9%
Company
Private sector
40.0% 38.1%
Public sector
Percent

30.0%

20.0% 18.7%
17.3%

12.7%
10.8%
10.0%
4.8%
2.9%
1.6%2.2% 2.2%

0.0%
0 1-10 days 11-20 days 21-30 days 31-40 days 41-50 days 61 days or
off for off for off for off for off for more off for
holidays holidays holidays holidays holidays holidays
Annual leave categories

54
Work load and responsibility for things and persons

Subjective quantitative work load

Subjective quantitative work load was determined by asking respondents to indicate the extent to which
particular statements reflects the time demands of their work. A 5-point Likert scale was used with
potential responses ranging from very little (1) to very great (5). The mean values for each item are
shown below. Items were summed and a mean score obtained where a higher score is indicative of an
increased perception of overload. The average score for the scale was 3.48 (SD=.457) indicating
respondents believed they had a moderately high quantitative work load. The Cronbachs alpha
coefficient for the scale was .71 indicating an acceptable level of internal consistency reliability.

Table 9 Subjective quantitative work load mean item scores

Mean Std.
Value Deviation
The number of projects and/or assignments you have. 3.85 0.720
The amount of time you spend in meetings. 3.07 0.912
The amount of time you have to undertake the work required of you. 3.25 0.823
The number of conflicting demands you have. 3.52 0.848
The work load, the amount of things that need to be done. 3.94 0.678
The time to think and contemplate. 2.51 0.865
The quantity of work you are expected to do. 3.94 0.670
The extent to which you feel you never have any time. 3.39 0.925
The number of phone calls and office visits you have during the day. 3.40 0.974

Responsibilities for things and people

Work responsibility was determined by two scales to measure both the responsibilities for things and
people. Respondents were asked to indicate the extent to which some statements reflect their level of
responsibility at work. A 5-point Likert scale was used ranging from very little (1) to very great (5).
The mean values for each item for the two scales are shown below. Items were summed and a mean
score obtained where a higher score is indicative of an increased perception of responsibility.

The average score for the responsibility for things scale was 3.44 (SD= 0.721) indicating respondents
believed they had a moderately high level of responsibility. The Cronbachs alpha coefficient for the
scale was .70 indicating an acceptable level of internal consistency reliability.

Table 10 Responsibility for things mean item scores

Mean Std.
Value Deviation
The responsibility you have for initiating assignments and projects. 3.49 0.953
The responsibility you have for budgets and expenditures. 3.38 1.149
The responsibility you have for carrying out assignments and
4.07 0.683
projects.
The responsibility you have for equipment and facilities. 2.82 1.114

The average score for the responsibility for people scale was 3.431(SD=1.0391) indicating
respondents believed they had a moderately high level of responsibility. The Cronbachs alpha
coefficient for the scale was .78 indicating an acceptable level of internal consistency reliability.

55
Table 11 Responsibility for people mean item scores
Mean Std.
Value Deviation
The responsibility you have for the safety of others. 3.41 1.283
The responsibility you have for the work of others. 3.46 1.065
The responsibility you have for the future careers of others. 2.82 1.111

Support at work
Support at work was measured by considering organisational support, supervisor support and co-
worker support.

Organisational support
Organisational support was determined by asking respondents about the extent to which they agreed or
disagreed with 16 statement relating to their organisation. A 7-point Likert scale was used ranging from
strongly disagree (1) to strongly agree (7). The average values for each item are presented in the table
below. Item scores were summed and a mean obtained. A high mean score is taken to indicate that the
person perceived they had high organisational support.

Table 12 Organisational support mean item scores

Mean Std.
Value Deviation
The organisation values my contribution to its well-being. 4.73 1.501
If the organisation could hire someone to replace me at a lower salary
3.81 1.686
it would do so.
The organisation fails to appreciate any extra effort from me. 4.55 1.662
The organisation strongly considers my goals and values. 3.93 1.557
The organisation would ignore any complaint from me. 3.31 1.527
The organisation disregards my best interests when it makes
3.85 1.503
decisions that affect me.
Help is available from the organisation when I have a problem. 5.17 1.223
The organisation really cares about my well-being. 4.18 1.499
Even if I did the best job possible, the organisation would fail to notice. 3.65 1.626
The organisation is willing to help me when I need a special favour. 4.55 1.371
The organisation cares about my general satisfaction at work. 4.37 1.440
If given the opportunity, the organisation would take advantage of me. 4.25 1.551
The organisation shows very little concern for me. 3.52 1.483
The organisation cares about my opinions. 4.46 1.382
The organisation takes pride in my accomplishments at work. 4.29 1.414
The organisation tries to make my job as interesting as possible. 3.93 1.398

The Cronbachs alpha coefficient for the scale was 0.94 indicating a high level of internal consistency
reliability. The overall average score of 4.29.

Supervisor support
Supervisor support was measured using a 13 item scale. Respondents were asked to read a list of
behavioural attributes and indicate the frequency with which their immediate supervisor behaved
towards them. Immediate supervisor was defined as the person who most often officially assesses your

56
job performance. A 5-point Likert frequency scale was used ranging from not at all (1) to all of the
time (5).

As the scale consisted of items from two separate scales a factor analysis was carried out determine the
modified scales factorial structure. One item double loaded on factors 1 and 2 so was removed from
the analysis because it could not be reliably interpreted. Table 13 shows the factor loadings for the
remaining items in the scale. This three-factor solution explained 69.2% of the variance in our sample.
These factors were labelled emotional support; practical support and criticism; and had an
average score of 3.66 (SD=.833), 3.25 (SD=.996) and 2.28 (SD=.457), respectively. These factors will
be considered separately in further analyses. The Cronbachs alpha coefficients were 0.92, 0.86 and 0.66
respectively indicating an acceptable level of internal consistency reliability.

Table 13 Factor analysis for supervisor support

Factor 1 Factor 2
Factor 3
Emotional Practical
Criticism
Support Support
Is concerned about me as a person. .838
Feels each of us is important as an individual. .833
Would help me to figure out how to solve a problem. .762
Is helpful to me when I have a family or personal emergency. .740
Is helpful to me when I have a routine family or personal
.637
matter to attend to.
Is understanding when I have personal or family problems
.625
which interfere with my work.
Appears to know a lot about company policies that help
.623
employees manage their family responsibilities.

Would switch schedules (hours, overtime hours, vacation) to

.864
accommodate my family responsibilities.
Is willing to juggle tasks or duties to accommodate my family
. .837
responsibilities.

Is critical of my efforts to combine work and family. .825

Shows resentment of my needs as a working parent. .689
Would hold my family responsibilities against me. .645
NB: Boxes denote factors

Table 14 Supervisor support mean scores

Mean Std.
Value Deviation
Emotional support 3.66 0.833
Practical support 3.25 0.996
Criticism 2.28 0.457

57
Co-worker support
Co-worker support was measured using two separate scales. The first scale focus is on the level of trust
or faith respondents perceive they have with their peers at work. This taps into the practical support
provided by co-workers. The second scale taps into the emotional support provided by co-workers.
Co-workers were defined as the people with whom you have the most contact in the company
(excluding immediate supervisor).

Co-worker support (practical) was measured using a 7-point Likert scale indicating the respondents
agreement with statements about their work group, and the extent to which they agreed or disagreed
with the statements in relation to their co-workers. The mean value for each of the items is provided in
Table 15.

Table 15 Co-worker support (practical) mean item scores

Mean Std.
Value Deviation
If I got into difficulties at work I know my workmates would try and
5.54 1.229
help me out.
I can trust the people I work with to lend me a hand if I need it. 5.55 1.292
Most of my workmates can be relied upon to do as they say they will
5.37 1.365
do.

All items were scored in the same direction, and summed. The mean score for the scale was 16.46
(SD=3.585) which ranges from a possible 3 to 21 indicating some agreement with the statements.
The Cronbachs alpha coefficient was .91 indicating an acceptable level of internal consistency
reliability.

Co-worker support (emotional) was measured using a 5-point Likert scale indicating the respondents
agreement to statements about the behaviour of their co-workers. The scale ranged from to a very
little extent (1) to to a very great extent (5). A higher mean score was indicative of a high degree of
perceived support from co-workers. The mean score for each item is presented in Table 16. The mean
for the scale was 3.67 (SD=.7269) and the Cronbachs alpha coefficient was .86 indicating an acceptable
level of internal consistency reliability.
Table 16 Co-worker support (emotional) mean item scores
Mean Std.
Value Deviation
How friendly or easy to approach are the persons in your work
3.87 0.821
group?
When you talk with persons in your work group to what extent do
3.82 0.727
they pay attention to what you're saying?
To what extent are persons in your work group willing to listen to
3.33 0.919
your problems?

Feelings about work

Respondents were asked to complete a number of scales which investigated their feelings about their
work situation. These included organisational commitment, turnover intent, job satisfaction and
burnout.

Organisational commitment
Organisational commitment was measured using an 18-item scale and respondents were asked to
indicate their agreement or disagreement on a 7-point scale ranging from strongly disagree (1) to
strongly agree (7). As commitment is multi-dimensional, a factor-analysis was carried out to confirm
the factorial structure in our sample. This was found to be in accordance with work by Allen and Meyer
(1990) and is presented in Table 17.

58
The first factor, affective commitment to the organization consisted of items numbered: 1, 4, 6, 7, 12,
and 14 (items 4, 6 and 14 were reverse coded). Continuance commitment consisted of items numbered
3, 8, 10, 11, 13 and 18, with item number 18 being reverse coded. The final factor (items 2, 5, 9, 15, 16
and 17) normative commitment contained only one reverse coded item (item 2).
Table 17 Factor analysis for organisational commitment
Factor 1 Factor 2 Factor 3
Affective Continuance Normative
Commitment Commitment Commitment
I do not feel like part of the family at my organisation.* .815
I do not feel a strong sense of "belonging" to my organisation.* .793
This organisation has a great deal of personal meaning for me. .776
This organisation deserves my loyalty. .717
I would be very happy to spend the rest of my career with this
.690
organisation.
I do not feel "emotionally attached" to this organisation.* .659
I owe a great deal to my organisation. .621
I feel that I have too few options to consider leaving this
.841
organisation.
One of the few negative consequences of leaving this organisation
.779
would be the scarcity of available alternatives. *
Too much of my life would be disrupted if I decided I wanted to
.716
leave my organisation now.
Right now, staying with my organisation is a matter of necessity as
.696
much as desire.
It would be very hard for me to leave my organisation right now,
.668
even if I wanted to.
Even if it were to my advantage, I do not feel it would be right to
.759
leave my organisation now.
I would feel guilty if I left my organisation now. .692
I would not leave my organisation right now because I have a sense
.672
of obligation to the people in it.
I really feel as if this organisations problems are my own. .553
If I had not already put so much of myself into this organisation, I
.548
might consider working elsewhere.*
I do not feel any obligation to remain with my current employer.* .426
NB: Boxes denote factors
* indicate reverse scoring

Mean scores for each of the three commitment scales were obtained and are presented in Table 18. A
high score indicated a high degree of commitment to the organisation. The Cronbachs alpha
coefficients were .86, .70 and .80 for affective, continuance and normative commitments respectively
indicating an acceptable level of internal consistency reliability.
Table 18 Organisational commitment mean scores
Mean Std.
Value Deviation
Affective commitment 4.34 1.376
Continuance commitment 4.11 1.141
Normative commitment 3.83 1.212

59
Turnover intent
Turnover intent was measured using a 2-item scale in which respondents were asked to indicate the
extent to which they agreed or disagreed with statements relating to thoughts about quitting and
looking for alternative work. A 7-point Likert scale was used ranging from strongly disagree (1) to
strongly agree (7). The mean value for each item is provided in the table below. The average score for
the scale was 3.398 (SD=1.814). The Cronbachs alpha coefficient for the scale was .78 indicating an
acceptable level of internal consistency reliability.

Table 19 Turnover intent mean item scores

Mean Std.
Value Deviation
I often think about quitting. 3.49 1.970
I will probably look for a new job in the next year. 3.30 2.030

Job satisfaction
Job satisfaction was measured using a scale developed at the University of Michigan (Taylor & Bowers,
1972) which consisted of three sub-scales: intrinsic rewards (items 1, 2, and 3); extrinsic rewards (items
4, 5, and 6); and social rewards satisfaction (items 7, 8, and 9). The 7-point Likert scale asked
respondents to indicate how satisfied they were with different aspects of their job. The scale ranged
from very dissatisfied (1) to very satisfied. A higher score is indicative of a greater level of
satisfaction. The mean values for each item is shown in the table below.

Table 20 Job satisfaction mean item scores

Mean Std.
Value Deviation
How satisfied are you with the chances you have to learn new things? 5.22 1.412
How satisfied are you with the chances you have to accomplish something worthwhile? 5.13 1.450
How satisfied are you with the chances you have to do something that makes you feel
4.92 1.499
good about yourself as a person?
How satisfied are you with the amount of pay you get? 4.19 1.694
How satisfied are you with the fringe benefits you receive? 3.88 1.732
How satisfied are you with the amount of job security you have? 5.33 1.563
How satisfied are you with the way you are treated by the people you work with? 5.28 1.369
How satisfied are you with the respect you receive from the people you work with? 5.11 1.397
How satisfied are you with the friendliness of the people you work with? 5.57 1.265

The Cronbachs alpha coefficient for the extrinsic reward items was .55, so item 6 was removed to
increase the scale reliability. The Cronbachs alpha coefficient was recalculated and at .78 indicated an
acceptable level of internal consistency reliability. The mean value for each scale is provided in Table
21. The Cronbachs alpha coefficient were .90, .70 and .92, respectively.

Table 21 Job satisfaction mean scores

Mean Std.
Value Deviation
Intrinsic rewards satisfaction 5.09 1.331
Extrinsic rewards satisfaction 4.03 1.501
Social rewards satisfaction 5.32 1.245

60
Burnout
Burnout was measured using the Maslach Burnout Inventory General Survey. Mean values and Standard
deviations for each item are presented in Table 22. Respondents were asked to read statements of job-
related feelings and decide if they ever felt this way about their job. Frequency of the feeling is indicated
on a 7-point Likert scale ranging from 0 (never) to 6 (every day). The 16-item inventory comprises three
sub-scales assessing emotional exhaustion, cynicism and professional efficacy. The items for the third
dimension of burnout are framed in positive terms, and thus a low score reflects a low sense of
professional efficacy.

Table 22 Burnout mean item scores

Mean Std.
Value Deviation
I feel emotionally drained from my work. 3.35 1.648
I have become less interested in my work since I started this job. 2.07 1.706
I have accomplished many worthwhile things in this job. 3.21 1.528
I feel tired when I get up in the morning and have to face another day on the
2.90 1.862
job.
Working all day is really a strain for me. 2.23 1.798
I can effectively solve the problems that arise in my work. 5.05 1.141
I feel burned out from my work. 2.53 1.768
I feel I am making an effective contribution to what my organisation does. 4.27 1.590
I have become less enthusiastic about my job. 2.49 1.750
In my opinion, I am good at my job. 4.76 1.223
I feel exhilarated when I accomplish many worthwhile things in this job. 3.78 1.551
I feel used up at the end of the work day. 2.88 1.858
I just want to do my job and not be bothered. 2.57 1.894
I have become more cynical about whether my work contributes anything. 2.30 1.960
I doubt the significance of my work. 1.84 1.875
At my work, I feel confident that I am effective at getting things done. 4.75 1.187

As burnout is a multi-dimensional, a factor-analysis was carried out to determine the factorial structure
in our sample. This factor analysis was undertaken as a precursor to further analysis because, where
variables are multi-dimensional, the antecedents and outcomes of each dimension may differ. The
results of this factor analysis are presented Table 23.

61
Table 23 Factor analysis for burnout
Factor 1 Factor 2
Factor 3
Emotional Professional
Cynicism
exhaustion efficacy
Working all day is really a strain for me. .849
I feel burned out from my work. .834
I feel used up at the end of the work day. .795
I feel tired when I get up in the morning and have to
.766
face another day on the job.
I feel emotionally drained from my work. .735
At my work, I feel confident that I am effective at
.803
getting things done.
In my opinion, I am good at my job. .776
I feel I am making an effective contribution to what my
.763
organisation does.
I feel exhilarated when I accomplish many worthwhile
.692
things in this job.
I can effectively solve the problems that arise in my
.658
work.
I have accomplished many worthwhile things in this
.637
job.
I doubt the significance of my work. .825
I have become more cynical about whether my work
.799
contributes anything.
I have become less enthusiastic about my job. .687
I have become less interested in my work since I
.625
started this job.
I just want to do my job and not be bothered. .545
NB: Boxes denote factors

In our sample, a three-factor model of burnout explained 65.1% of the variance. Item loadings were
consistent with the Maslach Burnout Inventory. Internal reliability coefficients for emotional
exhaustion, professional efficacy and cynicism were .90, .83 and .87 respectively, indicating a high level
of internal consistency reliability. The three factors of emotional exhaustion, professional efficacy
and cynicism were therefore retained and considered separately in further analyses. The average
values, on a scale of 0 for never to 6 for every day, are provided Table 24.

Table 24 Burnout mean scores

Mean Std.
Value Deviation
Emotional exhaustion 2.78 1.522
Professional efficacy 4.30 1.029
Cynicism 2.25 1.497

62
Relationship with spouse/partner
Aspects of the relationship between respondents and their spouse/partner were measured using four
scales: relationship sociability; companionship with partner; relationship tension; and social
undermining.

Relationship sociability and companionship with partner,

For these first two scales participants were asked to indicate how frequently they and their partner had
undertaken various things together in the past few weeks. The third scale asked participants to indicate
how frequently they and their partner had disagreed or had differences of opinions about specific issues
in the past few weeks. A 5-point Likert scale was used ranging from not at all (0) to very frequently
(4). A summed score for each of the scales was then calculated.

The mean score for the relationship sociability scale was 11.34 (SD=3.673). The Cronbachs alpha
coefficient for the scale was .84 indicating an acceptable level of internal consistency reliability. Individual
scores for each item are provided in the table below. The mean item score for the scale was 2.82
(SD=.917).

Table 25 Relationship sociability mean item scores

Mean Std.
Value Deviation
Visited friends together. 3.06 1.011
Entertained friends in your home. 2.76 1.052
Ate out together in a restaurant. 2.79 1.208
Gone out together to a cinema, to play sport or other
2.70 1.195
entertainment.

The mean score for the companionship with partner scale was 17.68 (SD=3.928). The Cronbach
alpha coefficient for the scale was .86 indicating an acceptable level of internal consistency reliability.
Individual scores for each item are provided in the table below. The mean item score for the scale was
3.54 (SD=.785).

Table 26 Companionship with partner mean item scores

Mean Std.
Value Deviation
Had a good laugh together or shared a joke. 3.76 .931
Been affectionate towards one another. 3.91 .910
Spent an evening chatting with one another. 3.49 1.043
Did something the other particularly appreciated. 3.47 .794
Taken a drive or walk for pleasure. 3.06 1.203

Relationship tension
In this scale participants were asked to indicate how frequently they and their partner had disagreed or
had differences of opinions about specific issues in the past few weeks. Again a 5-point Likert scale was
used ranging from not at all (0) to very frequently (4). A summed score for each of the scales was
then calculated. The mean score for the companionship with partner scale was 21.44 (SD=6.788). The
Cronbach alpha coefficient for the scale was .85 indicating an acceptable level of internal consistency
reliability. Individual scores for each item are provided in Table 27. The mean item score for the scale
was 2.39 (SD=.753)

63
Table 27 Relationship tension mean item scores
Mean Std.
Value Deviation
Being tired. 2.76 1.144
Irritating personal habits. 2.42 0.937
Household expenses. 2.39 1.169
Being away from home. 2.66 1.302
How to spend leisure time. 2.64 1.087
Time spent with friends. 2.26 1.015
Your partners job. 2.04 1.145
In-laws. 2.06 1.151
Not showing love. 2.25 1.077

Social undermining
Social undermining was measured by asking questions about the way their partner treated them.
A 3-point Likert scale ranging from never (1) to always (3) indicated the perceived level of social
undermining received. The mean values for each of the items is presented in Table 28. Once reversed
items were accounted for the whole scale was revered so that a higher score indicated a higher level of
perceived social undermining. Items were then summed. The mean value for the scale was 1.47
(SD=0.287) with an alpha reliability of .852.

Table 28 Social undermining mean item scores

Mean Std.
Value Deviation
Supports and reassures you when you need it. 2.64 0.521
Treats you with anger and aggression. 1.49 0.514
Shows that he/she cares about you as a person. 2.76 0.428
Complicates your life. 1.75 0.490
Provides you with useful information or advice when you need it. 2.41 0.518
Shows that he/she does not like you. 1.21 0.441
Listens to you when you need to talk about things that are very
2.67 0.499
important to you.
Makes you feel unwanted. 1.24 0.485
Tells you things that strengthen your self confidence. 2.32 0.609
Gets on your nerves. 1.67 0.471
Understands your way of thinking. 2.25 0.518
Criticizes you. 1.66 0.489
Gives you direct help, i.e. does something for you or gives you what
2.47 0.537
is necessary.
Insults you even when he/she does not want to. 1.27 0.473
Makes you feel that you can count on him/her. 2.75 0.488

Parenting issues

Parenting behaviours
Respondents were asked a number of questions regarding their parental behaviour. In particular they
were asked to read a list of parenting behaviours and indicate how frequently or infrequently (from
never (1) to always (5)) they behaved in that way towards their child(ren). They were specifically

64
asked to relate their answers to their dependent children, i.e. those under 18 years. Individual scores for
each item are provided in the table below.

Table 29 Parenting behaviours mean item scores

Mean Std.
Value Deviation
I encourage my child to talk about the childs troubles. 3.43 1.207
I show sympathy when my child is hurt or frustrated. 4.08 0.802
I explode in anger towards my child. 1.85 0.628
I am aware of problems or concerns about my child in school. 3.24 1.216
I am responsive to my childs feelings or needs. 3.92 0.774
I yell or shout when my child misbehaves. 2.20 0.867
I give comfort and understanding when my child is upset. 4.14 0.877
I express affection by hugging, kissing and holding my child. 4.18 0.960
I argue with my child. 2.08 0.806
I have warm and intimate times together with my child. 3.73 1.017
I give praise when my child is good. 4.33 0.728
I am easy going and relaxed with my child. 3.99 0.734
I tell my child I appreciate what the child tries or accomplishes. 4.13 0.892
I show patience with my child. 3.76 0.874
I know the names of my childs friends. 3.56 1.024
I joke and play with my child. 4.01 0.863

The sixteen items were analysed, and the results suggested two main factors: nurture and support
child (items 1, 2, 4, 5, 7, 8, 10, 11, 12, 13, 14, 15, and 16) with a Cronbachs alpha of .87; and verbal
anger towards child (items 3, 6, and 9), with a Cronbachs alpha of .76. The scores for each factor were
summed and a mean obtained. The mean score for nurture and support child was 3.86 (SD=0.653)
and verbal anger towards child was 2.04 (SD=0.636).

Table 30 Parental attitudes mean scores

Mean Std.
Value Deviation
Nurture and support child 3.86 0.653

Verbal anger towards child 2.04 0.636

Support with dependants

Level of support with children was measured with a single item. Possible responses ranged from I/we
have no family or friends who could look after the children at 1 to I/we can always get family or
friends to look after the children at 1. A higher score indicated a higher level of perceived support with
children.

The average score was 2.64 (SD=0.933). Individual item results are shown in the Table 31.

65
Table 31 Support with dependants
Number %
I/we can always get family or friends to look after the children. 18 18.9
If I/we plan well ahead I/we can always get family or friends to
37 38.9
look after the children.
I/we can call on family/friends in an emergency situation. 28 29.5
I/we have no family or friends who could look after the children. 12 12.6

Work-life balance

Work-family conflict
Participants were asked to read 10 statements about their experiences at work, and outside of work, and
to indicate the extent to which they agreed or disagreed with them. The items were scored on a 7-point
scale ranging from strongly disagree (1) to strongly agree (7). A mean score for work-to-family
conflict was obtained by summing items 1, 2, 3, 4 and 5 and obtaining an average. Similarly, family-to-
work conflict was determined by obtaining a mean score from items 6, 7, 8, 9 and 10.

Table 32 Work-to-family conflict mean item scores

Mean Std.
Value Deviation
The demands of my work interfere with my home and family life. 4.88 1.495
Because of my job, I can't involve myself as much as I would like in
4.27 1.648
maintaining close relations with my family or spouse/partner.
Things I want to do at home do not get done because of the
4.55 1.648
demands my job puts on me.
I often have to miss important family activities because of my job. 3.48 1.714
There is a conflict between my job and the commitments and
3.98 1.700
responsibilities I have to my family or spouse/partner.

The mean score for each work-to-family conflict item is shown in Table 32. The overall mean score for
work-to-family conflict was 4.33 (SD=1.873). The Cronbach alpha coefficient for the scale was .91
indicating an acceptable level of internal consistency reliability.

The mean score for each family-to-work conflict item is shown in Table 33. The overall mean score for
work-to-family conflict was 2.79 (SD=1.121). The Cronbach alpha coefficient for the scale was .80
indicating an acceptable level of internal consistency reliability.

Table 33 Family-to-work conflict mean item scores

Mean Std.
Value Deviation
The demands of my family or spouse/partner interfere with work-
3.27 1.532
related activities.
I sometimes have to miss work so that family responsibilities are
3.27 1.728
met.
Things I want to do at work don't get done because of the demands
2.54 1.353
of my family or spouse/partner.
My home life interferes with my responsibilities at work such as
getting to work on time, accomplishing daily tasks, and working 2.71 1.576
overtime.
My co-workers and peers at work dislike how often I am
2.17 1.335
preoccupied with my family life.

66
Family management
Family management was measured via a 21-item scale. Participants were asked to read a list of activities
and to indicate the extent to which they believed it was easy or difficult to perform them in their
present job. A higher score indicated a greater degree of difficulty requiring more family management
in order to undertake the activity. The first eight items were for parents, and those without children
responded with not applicable. The next 13 items were relevant to all participants. It was decided for
those parents who had mistakenly responded not applicable for the first 8 items that their score
would be replaced with the mean score of the parent sample. For those who were not parents and had
answered not applicable their score was excluded from the analysis.

In our sample, a forced three-factor model of family management explained 60% of the variance. A
varimax rotation was used. The three factors were named care for children, leisure time and work
flexibility. Some items, as indicated below, were removed as they did not load highly enough on any
one factor.

Table 34 Factor analysis for family management

Factor 1 Factor 2 Factor 3
Care of Leisure Work
children time flexibility
To make alternate child care arrangements when necessary. .773
To take your children to or from a child care setting or school. .729
To go to school events for your children. .714
To stay at home with a sick child. .674
To take your children to health care appointments. .624
To be home when your children get home from school. .570
To be home for services or deliveries (e.g. gas, appliances). .567 .338 .433
To go to health care appointments. .537 .352 .458
To spend fun or educational time with the family. .837
To visit or help neighbours or friends. .786
To participate in community activities. .698
To have meals with the family. .694
To have relaxed, pleasant times with your children. .691
To make arrangements for children during summer vacation. .478 .484
To go to work a little later than usual if you need to. .666
To avoid the rush hour. .658
To adjust your work hours to the needs of other family
.608
members
To take care of household chores. . .569
To make telephone calls for appointments or services. .543
To go on errands (e.g., post office, car service). .448 .347 .526
To go shopping (e.g. groceries, clothes). .477 .398 .489
NB: Boxes denote factors

The items in each factor were summed and a mean obtained. These are provided in Table 35. The
internal reliability coefficients for care of children, leisure time and work flexibility were .85, .89
and .78, respectively.

67
Table 35 Family management mean scores
Mean Std.
N
Value Deviation
Care for children 109 3.60 .760
Leisure time 201 2.86 .861
Work flexibility 201 2.78 .793

Work-life control
Respondents were asked to indicate the degree of choice (from very little (1) to very great (5)) they
had over issues relating to the work-life situation. Again non parents were given an option of not
applicable for questions relating to parenting issues. Those responses were excluded from the analysis.

In our sample, a forced two-factor model of work-life control explained 50% of the variance using a
varimax rotation. The two factors were named general work-life control and work-family control.
Five items, as indicated below, were excluded on the basis of factor loadings. A mean control score was
obtained for the final two factors.
Table 36 Factor analysis for work-life control
Factor 1 Factor 2
General Work-
work-life family
control control
In general, how much control do you have over the way you balance working
.858
and other responsibilities in your life?
In general, how much control do you have over the way you balance working
.834
and other interests in your life?
How much control do you have over when you can take a few hours off? .796
How much choice do you have over when you take vacations or days off? .783
How much choice do you have over when you begin and end each work day or
.757
each work week?
In general, how much control do you have over the way you balance working
.709
and parenting?
How much choice do you have in arranging part-time employment? .381
How much choice do you have over the amount and timing of work you must
.374 .362
do at home in order to meet your employment demands?
How much choice do you have in obtaining adult supervision for your
.836
child/children before or after school?
How much choice do you have over the amount and quality of day care
.823
available for your child/children?
How much choice do you have over the amount and quality of care available
.747
for a sick child?
How much choice do you have in making unanticipated child-care
.698
arrangements?
How much choice do you have over the amount and quality of day care
.644
available for a dependent parent or other relative?
How much choice do you have over the amount you pay for dependent care? .499
To what extent can you choose to do some of your work at home instead of
.455
your usual place of employment?
To what extent are you expected to limit the number of times you make or
receive personal phone calls while you work?
NB: Boxes denote factors

68
The items in each factor were summed and a mean obtained these are provided in Table 37. These
factors will be considered separately in further analyses. The internal reliability coefficients for general
work-life control and work-family control were .88 and .84, respectively.

Table 37 Work-life control mean scores

Mean Std.
N
Value Deviation
General work-life control 202 3.15 .814
Work-family control 96 2.65 .780

Work-life benefits
A list of 21 work-life benefits were presented and respondents were asked to indicate the usefulness of
each benefit to them at the present time. A 5-point Likert type scale was used ranging from not at all
useful (1) to very useful (5). The mean scores for each item are provided in the Table 38.

Table 38 Work-life benefits mean item scores

Mean Std.
Value Deviation
Providing child care facilities. 1.80 1.320
Ensuring people take their annual leave regularly. 2.67 1.227
Providing an information and referral service to assist with care of aged parents. 1.67 1.041
Providing an information and referral service to assist with care of dependent children. 1.67 1.085
Providing assistance with child care costs (i.e. salary sacrificing). 1.87 1.402
Providing emergency care for a child or other dependants. 2.00 1.407
Providing a wellness programme. 2.92 1.223
Reimbursing the costs of courses and further study. 3.62 1.370
Providing more flexible work hours. 3.63 1.190
Offering extended part time options for return to work after the birth or adoption of a child. 1.82 1.366
Providing a fitness programme (e.g. paying for sports club membership). 3.41 1.412
Introducing job sharing. 2.13 1.299
Provision of a legal information service. 2.67 1.286
Allowing for special family leave, eg to care for a sick dependant. 2.87 1.504
Offering permanent part-time work options. 2.35 1.471
Providing vacation care for children during school holidays. 1.80 1.344
Offering temporary part-time work options during family crises. 2.57 1.464
Providing scholarships for employees children 2.48 1.695
Increasing flexibility in work location (working from home/telecommuting). 3.31 1.365
Providing an employee assistance programme for employees with family problems. 2.60 1.364
Offering extended parental leave after the birth or adoption of a child. 2.03 1.502

A factor analysis with a varimax rotation was undertaken which explained 61% of variance. See Table
39. The following sub-scales were obtained: childcare support (items 1, 4, 5, 6, 16, 18, and 21); alternate
work arrangements (items 10, 12, 15, and 19); crisis assistance/support (items 2, 3, 13, 14, 17, and 20);
and wellness and personal development (items 7, 8, 9, and 11).

69
Table 39 Factor analysis for work-life benefits
Factor 1 Factor 2 Factor 3 Factor 4
Childcare Alternative Crisis Wellness
support work assistance and personal
arrangements /support development
Providing child care facilities. .893
Providing emergency care for a child or other dependants. .889
Providing assistance with child care costs (i.e. salary sacrificing). .887
Providing vacation care for children during school holidays. .821
Providing an information and referral service to assist with care
.798
of dependent children.
Offering extended parental leave after the birth or adoption of a
.557
child.
Providing scholarships for employees children .520
Offering permanent part-time work options. .799
Introducing job sharing. .745
Offering extended part-time options for return to work after the
.667
birth or adoption of a child.
Increasing flexibility in work location (working from
.549
home/telecommuting).
Provision of an information and referral service to assist with
.765
care of aged parents.
Providing of a legal information service. .644
Offering temporary part-time work options during family crises. .630
Allowing for special family leave, eg to care for a sick dependant. .615
Providing an employee assistance programme for employees
.599
with family problems.
Ensuring people take their annual leave regularly. .568
Providing a fitness programme (e.g. paying for sports club
.746
membership).
Providing a wellness programme. .658
Reimbursing the costs of courses and further study. .646
Providing more flexible work hours. .567
NB: Boxes denote factors

The mean values for each factor is shown in the table below. The alpha reliability for childcare
support, alternative work arrangements, crisis assistance/support and wellness and personal
development was .93, .77, .79 and .87, respectively.

Table 40 Work-life benefit preferences mean scores

Mean Std.
Value Deviation
Childcare support 1.95 1.116
Alternative work arrangements 2.41 1.059
Crisis assistance/support 2.51 0.932
Wellness and personal development 3.39 0.919

In particular, it is interesting to note the change in preferences for work-life benefits for various age
groups and family structures. As can be seen, child care support initiatives become more popular in the
30 to 49 age groups, and wellness and personal development initiatives decrease in preference with age.

70
Table 41 Work-life benefit preferences mean score for various age cohorts
Mean Value
20-29 years 30-39 years 40-49 years 50-59 years 60 years +
Childcare support 1.76 2.16 2.22 1.62 1.07
Alternative work arrangements 2.49 2.51 2.06 2.57 1.83
Crisis assistance/support 2.44 2.39 2.55 2.82 2.05
Wellness and personal development 3.68 3.49 3.37 3.09 2.82

Figure 33 Work-life benefit preferences for various age cohorts

It can be seen from Table 42 and Figure 34 that child care initiatives, alternative work arrangements and
crisis assistance/support are most favoured by those with dependent children. Couples without
children and single persons are more interested in work-life initiatives which focus on wellness and
personal development.

Table 42 Work-life benefit preferences mean score for different family structures
Mean Value
Couple Couple non Couple
Single Single
dependent dependent without
parent person
children children children
Childcare support 2.69 1.31 1.79 1.50 1.33
Alternative work arrangements 2.45 2.34 2.30 2.43 2.37
Crisis assistance/support 2.73 2.39 2.14 2.30 2.40
Wellness and personal development 3.34 3.06 3.17 3.54 3.63

71
Figure 34 Work-life benefit preferences mean score for different family structures

While it should be noted that the female sample size was small some gender differences in work-life
benefits were apparent. However the results, which are similar to the single and couple without
children results above, do appear to reflect the fact the female sample is a younger cohort than their
male counterparts. See Figure 35.

Figure 35 Work-life benefit preferences mean score by gender

72
Comparison between the public and private sectors

In order to compare employees responses on key variables between public and private sector
organisations, we conducted a series of independent-samples t-tests to identify significant differences
The differences in the sample means are expressed as coefficients (t) in t-tests and the significance is
indicated by the probability of achieving this result by chance (p). A probability of equal to or less than
.05 was deemed to be significant.

Work load and responsibility

Work load
Figure 36.depicts the mean scores for the private and public sector employees for the variable
subjective qualitative workload. Public sector respondents indicated slightly higher levels of subjective
quantitative workload than private sector employees. However, the t-test indicated that the difference
between sectors was not statistically significant (t (200) = -1.024, p = .31).

Figure 36 Comparison of mean scores of subjective qualitative workload for public and
private sector

4.00
3.4321 3.5032

3.00
Mean

2.00

1.00

0.00
Private sector Public sector
Company

Responsibility
Figure 37 depicts the mean scores for the private and public sector employees for two dimensions of
work responsibility (responsibility for things and responsibility for persons). Private sector
employees scored slightly higher for both of these dimensions than their public sector counterparts, but
the t-tests revealed that the differences between sectors were not statistically significant for
responsibility for things (t (200) = 1.531, p = .127) but was significant for responsibility for persons
(t (200) = 2.193, p = .029). This indicates that project based employees in the private sector consider
themselves to have a higher responsibility for other people than their counterparts in the public sector.

73
Figure 37 Comparison of mean scores for responsibility for things and persons for the
public and private sector

4.00
3.667
3.556
3.388 3.324

3.00
Responsiblity for
things
Mean

2.00 Responsiblity for

persons

1.00

0.00
Private sector Public sector
Company

Work hours
Figure 38 depicts the mean number of work-related hours spent away from the home (at work or
commuting) during the week prior to respondents completing the survey. Private sector employees
reported being absent from home due to work for a greater amount of time than public sector
employees (i.e. 57 hours versus 43 hours). The t-test revealed that this difference was highly statistically
significant (t (97) = -9.266, p = .000).

Figure 38 Comparison of mean weekly work hours for public and private sector

60 56.52

50
43.26
Mean (hours)

40

30

20

10

0
Private sector Public sector
Company

Support at work

Perceived Organisational Support (POS)

Figure 39 depicts the mean scores for the private and public sector employees for the variable (POS).
The analysis revealed that private sector respondents reported a slightly higher level of POS than their
public sector counterparts. However, the difference between sectors was not statistically significant
(t (200) = 0.618, p = .538).

74
Figure 39 Comparison of mean perceived organisational support for public and private
sector

5.00
4.3633 4.2625
4.00

3.00
Mean

2.00

1.00

0.00
Private sector Public sector
Company

Supervisor support
Figure 40 depicts the mean scores for the private and public sector employees for three dimensions of
supervisory support (emotional, practical and resentment/criticism). Public sector employees reported
slightly higher levels of support and a lower level of resentment/criticism than their private sector
counterparts. However, the t-tests indicated that the differences between sectors for two dimensions
were not statistically significant (emotional support, t (200) = -1.349, p = .179; and
resentment/criticism, t (200) = 0.568, p = .571). However, the difference between sectors for the
dimension of practical support from ones supervisor was statistically significance (t (200) = -2.579,
p = .011), indicating that supervisors in the public sector are systematically more likely to lend practical
assistance to subordinates with family obligations than supervisors in the private sector.

Figure 40 Comparison of the means of the three dimensions of supervisory support for
the public and private sector

4.00
3.713
3.543
3.365
2.98
3.00
Emotional support
Practical support
Mean

Criticism
2.00 1.85 1.785

1.00

0.00
Private sector Public sector
Company

Practical co-worker support

Figure 41 depicts the mean scores for the private and public sector employees for the variable of
practical co-worker support. Employees in the public sector reported slightly higher levels of

75
practical co-worker support than their private sector counterparts. However, the t-test indicated that
differences between the sectors were not statistically significant (t (200) = -1.068, p = .287).

Figure 41 Comparison of means for practical co-worker support for public and private
sector respondents

20.00

16.637
16.055
15.00
Mean

10.00

5.00

0.00
Private sector Public sector
Company

Affective co-worker support

Figure 42 depicts the mean scores for the private and public sector employees for the variable
affective co-worker support. This relates to emotional support provided by ones co-workers, as
opposed to practical assistance. Public sector employees reported higher levels of affective co-worker
support than their private sector counterparts. However, the t-test revealed that this difference was not
statistically significant (t (200) = -1.637, p = .103).

Figure 42 Public and private sector means for affective co-worker support

4.00 3.7251
3.5452

3.00
Mean

2.00

1.00

0.00
Private sector Public sector
Company

76
Feelings about work

Organisational commitment
Figure 43 shows the mean scores for the three dimensions of organisational commitment (normative,
affective and continuance) for public and private sector respondents. Public sector respondents scored
slightly higher than private sector employees in all dimensions of organisational commitment. However,
the t-tests indicated that the differences between sectors were not statistically significant for either
affective commitment (t (200) = -1.100, p = .273) or normative commitment (t (200) = -0.294,
p = .769). However, a highly significant difference was found between public and private sector
employees for the dimension of continuance commitment (t (200) = -4.131, p = .000). This is a concern
for public sector managers because continuance commitment is consistently linked to lower levels of
organisational performance (See Introduction).

Figure 43 Comparison of mean organisational commitment scores for public and

private sector respondents

5.00
4.413 4.325
4.183 Organisational
4.00 3.791 3.845 commitment
3.637
Affective
3.00
Mean

Continuance
Normative
2.00

1.00

0.00
Private sector Public sector
Company

Turnover Intention
Figure 44 depicts the mean scores for turnover intention for both the private and the public sector.
Private sector employees reported higher levels of turnover intention than public sector employees.
however, the t-test revealed that this difference was not statistically significant (t (199) = 1.678,
p = .095).

Figure 44 Comparison of public and private sector mean turnover intention scores

4.00 3.714

3.254
3.00
Mean

2.00

1.00

0.00
Private sector Public sector
Company

77
Job satisfaction
Figure 45 shows the mean scores for intrinsic, extrinsic and social rewards offered by work for both
public and private sector respondents. However, the t-tests revealed that the differences between
sectors in means for each type of job reward were not statistically significant (intrinsic rewards,
t (154) = 1.594, p = .113; extrinsic rewards, t (200) = 1.658, p = .099; and social rewards,
t (104) = 102.48, p = .201).

Figure 45 Comparison of public and private sector mean scores for job satisfaction

6.00
5.291 5.403
5.143 5.00
5.00
Job satisfaction
4.294
4.00 3.917 Intrinsic rewards
Extrinsic
Mean

3.00 rewards
Social rewards
2.00

1.00

0.00
Private sector Public sector
Company

Burnout
Figure 46 depicts the mean scores for both the public and private sector employees for each of the
three dimensions of burnout (i.e. emotional exhaustion, cynicism and professional efficacy). No
significant differences between the two sectors were identified for two dimensions of burnout (cynicism
t (200) = -1.856, p = .065; and emotional exhaustion t (200) = 1.780, p = .077). However a significant
statistical difference was found for the professional efficacy dimension (t (200) = 2.954, p = .004). Items
on this dimension of burnout are framed in positive terms indicating the public sector respondents
have a significantly lower sense of professional efficacy than their private sector colleagues.

The levels of burnout were relatively high in both sectors. Thus, there is cause for concern in both
organisations. For further information see Burnout analysis section in the report

Figure 46 Public and private sector mean burnout scores

5.00
4.61
4.16
4.00 Burnout

3.06
Emotional exhaustion
3.00 Cynicism
Mean

2.65
2.38 Professional efficacy
1.96
2.00

1.00

0.00
Private sector Public sector
Company

78
Spouse/partner relationship

Compatibility and sociability of partnerships (domestic partners)

Figure 47 depicts the mean scores for both the public and private sector respondents for partnership
compatibility and partnership sociability. The t-tests revealed no significant differences between
employees in the two sectors for either the compatibility scale (t (157) = -1.201, p =.231) or the
sociability scale (t (157) = 0.164, p = ..870).

Figure 47 Public and private sector mean scores for compatibility with partner,
relationship sociability and relationship tension

4.00
3.595
3.437

3.00 2.838 2.813

2.498
Relationship
2.329 sociability
Mean

2.00 Companionship
with partner
Relationship
1.00 tension

0.00
Private sector Public sector
Company

Partnership tensions
Figure 47 also depicts the mean scores for both the public and private sector respondents for the
partnership tensions scale. No significant differences between the respondents from the two sectors
were detected for this scale (t (157) = 1.342, p = .182).

Taken together, the results presented in the above figure suggest that the sector in which a person
works has little influence on the quality of their family relationships. This is perhaps unsurprising,
because it is likely that the relationship between the work context and family relationships will be
mediated by the quality of employees experiences at the work-family interface, such as work-family
conflict, stress, fatigue and burnout. These experiences are likely to vary within rather than between
organisations.

Work-life balance

Work-family conflict
Figure 48 depicts the mean scores for work-family conflict (in both directions) reported by public and
private sector respondents. A t-test revealed highly significant differences between the levels of work
interference with family (t (200) = 2.936, p = .004) An examination of the mean scores for this variable
revealed that private sector employees perceive significantly higher levels of work interference with
family than public sector employees. The mean scores for family interference with work were very
similar for both public and private sector respondents. As expected, the t-test for this variable failed to
reveal significant differences (t (200) = -0.216, p = .829).

79
Figure 48 Public and private sector mean scores for work-to-family and family-to-work
conflict scores

5.00 4.89

4.07
4.00 Work-family conflict
Family-work conflict
3.00 2.80
Mean

2.77

2.00

1.00

0.00
Private sector Public sector
Company

Family management
Figure 49 depicts the mean scores for both the public and private sectors on the three dimensions of
lifestyle management (leisure time with family/friends, work flexibility and care of children). An
examination of the means revealed that private sector employees scored more highly on each of these
dimensions. A t-test revealed all of these differences to be highly significant; care of children
t (107) = 4.467, p = .000; leisure time with family/friends t (199) = 6.011, p = .000; and work flexibility,
t (199) = 5.738, p = .000.

Figure 49 Public and private sector mean scores for family management scores

3.992
4.00

3.369 3.299 3.37

3.00 2.76 2.738 Care of children

Leisure time
Mean

2.00 Work flexibility

1.00

0.00
Private sector Public sector
Company

Control/flexibility - work and children

Figure 50 depicts the mean scores for both the private and public sectors on two dimensions of
control/flexibility (work and children). Private sector employees reported less flexibility with respect to
work. This difference was highly statistically significant t (200) = -6.702, p = .000. The difference
between sectors on the dimension of flexibility relating to children was not statistically significant
t(94) = -1.704, p = .092.

80
Figure 50 Public and private sector mean scores for the two dimensions of
control/flexibility (work and children)

4.00

3.259 General work-life control

3.00 2.754 Work-family control
2.651
2.469
Mean

2.00

1.00

0.00
Private sector Public sector
Company

Work-life benefit scale

Figure 51 depicts the mean scores for both sectors on the four subscales of Work-Life Benefit Scale
(childcare support, alternate work arrangements, crisis assistance/support and wellness/personal
development). A t-test failed to reveal significant differences between the two sectors on any of the
subscales: childcare support t (198) = 0.560, p = .578; alternate work arrangements t (199) = -0.728,
p = .467; and crisis assistance/support t (200) = -1.308, p = .192 and wellness/personal development
t (200) = 0.686, p = .493.

Figure 51 Public and private sector mean scores for the four dimensions of work-life
benefit.

81
Summary of public and private sector employee differences
In summary, private sector employees reported working significantly longer hours than public sector
employees. Private sector employees also reported having significantly less flexibility or control about
how they managed their work-life balance than public sector employees. Compared with public sector
employees, private sector employees found engaging in leisure activities with family and friends and
taking part in domestic and child rearing activities significantly more difficult. Private sector employees
reported significantly higher levels of work interference with family life than public sector employees.
Despite this, private sector employees did not report significantly higher levels of emotional exhaustion
or significantly lower levels of affective organisational commitment or job satisfaction than public
sector employees. Public sector employees also reported higher levels of continuance commitment than
private sector employees. This is a concern because the continuance commitment dimension of
organisational commitment has been consistently linked to lower levels of organisational performance.
Private sector employees express a significantly higher turnover intent than public sector employees.

In terms of supportiveness of the work environment, there were few significant differences between
public and private sector employees. However, employees in the public sector organisation reported
that they received higher levels of practical assistance from their immediate supervisor in balancing
work and family, than did employees in the private sector. This finding is important because
expressions of emotional support for employees work-family balance which are not backed up by
practical assistance have been linked to lower levels of employees affective commitment and
satisfaction.

82
Bi-variate correlations for employee data

Whole sample
Table 43 shows the bi-variate correlations between the major work-related variables measured in the
study. These results are presented under a series of sub-headings. Readers should note that significant
linkages are only described once. For example, if a linkage between burnout and work hours is
presented under the heading workload variables it is not re-stated under the heading Well-being (burnout).

Demographic variables
Employees age was significantly and positively related to organisational tenure, the total number of
children, days absent during the twelve months prior to the study (excluding holidays and leave days)
and continuance commitment. Organisational tenure was significantly and negatively related to the
amount of time employees devote to work on the weekend. Organisational tenure was also positively
related to the total number of children. Organisational tenure was significantly and positively related to
both affective and continuance organisational commitment. Employees total number of children was
also significantly and positively related to the continuance dimension of organisational commitment.

Workload variables
Average work hours were significantly and positively related to subjective qualitative workload, and
both dimensions of work-related responsibility, i.e. responsibility for things and responsibility for
persons. Work hours were also positively and significantly related to the continuance dimension of
organisational commitment and employees perceptions that work interferes with family.

Subjective qualitative workload was significantly and positively correlated with both dimensions of
work-related responsibility, i.e. responsibility for things and responsibility for persons. Subjective
qualitative workload was also significantly inversely correlated with Perceived Organisational Support
(POS) and extrinsic job satisfaction. Subjective qualitative workload was significantly and positively
correlated with employees perceptions that work interferes with their family life, and was also
significantly positively correlated with both the emotional exhaustion and cynicism dimensions of
burnout.

Work-responsibility (for things) was significantly and positively correlated with the other dimension of
work responsibility (for persons) in our sample. Work responsibility (for things) was also significantly
and positively correlated both the affective dimension of organisational commitment and intrinsic job
satisfaction, and significantly but inversely correlated with the continuance dimension of organisational
commitment. Work responsibility (for things) was significantly and positively correlated with employee
perceptions that their work interferes with family life. Work responsibility (for things) was significantly
and positively correlated to both the emotional exhaustion and professional efficacy dimensions of
burnout, suggesting that although heightened work responsibility is associated with higher emotional
exhaustion, it also contributes to employees sense of personal worth and effectiveness.

Work responsibility (for persons) was significantly and inversely correlated with POS. Work
responsibility (for persons) was also significantly inversely correlated with the continuance dimension of
organisational commitment. Work responsibility (for persons) was significantly and positively correlated
with employees perceptions that work interferes with family and their turnover intention. Work
responsibility (for persons) was also significantly and positively correlated to both the emotional
exhaustion and professional efficacy dimensions of burnout, lending further evidence to the
proposition that although heightened work responsibility is associated with higher emotional
exhaustion, it also contributes to employees sense of personal worth and effectiveness.

83
 Table 43 Bi-variate correlations between work variables measured in the study
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

1. Age 1

2. Tenure .572(**) 1

3. Work hours -.122 -.170(*) 1

4. Number of children .254(**) .310(**) .023 1

5. Subjective qualitative wk/load .014 .122 .238(**) .023 1

6. Work responsibility (things) .012 .115 .370(**) .081 .495(**) 1

7. Work responsibility (persons) -.013 -.021 .372(**) -.040 .452(**) .629(**) 1

8. Organizational support (POS) -.018 -.118 -.100 -.013 -.224(**) -.031 -.161(*) 1

9. Affective commitment .023 .178(*) -.089 .048 .028 .157(*) -.005 .614(**) 1

10. Continuance commitment .201(**) .329(**) -.287(**) .229(*) -.128 -.178(*) -.242(**) -.114 -.038 1

11. Normative commitment -.085 -.016 -.141(*) .051 -.041 .088 -.010 .541(**) .730(**) .022 1

12 Turnover intent -.055 -.073 .174(*) .003 .131 .055 .181(*) -.580(**) -.670(**) -.112 -.589(**) 1

13. Job Satisfaction - Intrinsic .017 -.075 -.006 .100 -.014 .200(**) .051 .667(**) .558(**) -.190(**) .489(**) -.468(**) 1

14. Job Satisfaction - Extrinsic -.030 -.115 -.027 .035 -.196(**) -.003 -.099 .442(**) .280(**) -.066 .285(**) -.299(**) .400(**) 1

15. Job Satisfaction - Social -.037 .015 -.108 .032 -.096 -.025 -.049 .483(**) .386(**) -.038 .223(**) -.294(**) .448(**) .215(**) 1

16. Burnout - emot exhaustion -.123 -.021 .262(**) -.059 .414(**) .160(*) .252(**) -.424(**) -.313(**) .096 -.232(**) .374(**) -.316(**) -.136 -.314(**) 1

17. Burnout - cynicism -.069 .057 .037 -.161 .205(**) -.028 .097 -.550(**) -.473(**) .179(*) -.382(**) .473(**) -.574(**) -.142(*) -.393(**) .656(**) 1

18. Burnout prof efficiacy .111 -.022 .134 -.025 .133 .263(**) .222(**) .304(**) .307(**) -.297(**) .282(**) -.205(**) .480(**) .237(**) .140(*) -.076 -.247(**) 1

19. Work-family conflict -.130 -.091 .343(**) -.105 .279(**) .317(**) .301(**) -.177(*) -.089 -.158(*) -.081 .147(*) -.012 -.154(*) -.145(*) .329(**) .109 .085 1

20. Family-work conflict -.031 .015 -.121 -.017 .078 .098 .037 .029 -.036 .002 -.054 .070 -.009 -.074 -.101 .041 -.053 -.158(*) .149(*)

** Correlation is significant at the 0.01 level (2-tailed).

* Correlation is significant at the 0.05 level (2-tailed).

84
Work-related attitudes
POS was positively and significantly correlated with both the affective and normative dimensions of
organisational commitment. POS was also inversely and significantly correlated with employees
turnover intention and their perception that their work interfered with their family life. POS was also
significantly inversely correlated with the emotional exhaustion and cynicism dimensions of burnout,
and positively correlated with employees sense of professional efficacy.

The affective dimension of organisational commitment was significantly and positively correlated with
the normative dimension of organisational commitment and all facets of job satisfaction as well as
being significantly inversely related to employees turnover intention. The affective dimension of
organisational commitment was also significantly and negatively correlated with the emotional
exhaustion and cynicism dimensions of burnout, and significantly and positively correlated with the
professional efficacy dimension of burnout.

The continuance dimension of organisational commitment was significantly inversely correlated with
employees perceptions that work interferes with family and intrinsic job satisfaction. The continuance
dimension of organisational commitment was also positively correlated with employees cynicism about
their work. Finally, the continuance dimension of organisational commitment was inversely correlated
with employees professional efficacy. The normative dimension of organisational commitment was
significantly inversely correlated with employees turnover intention. Normative organisational
commitment was also inversely correlated with the emotional exhaustion and cynicism dimensions of
burnout. Finally, the normative dimension of organisational commitment was significantly and
positively correlated with professional efficacy.

Turnover intention was positively and significantly correlated with employees perceptions that their
work interfered with their family life. Turnover intention was also significantly and positively correlated
with the emotional exhaustion and cynicism dimensions of burnout. Finally, turnover intention was
significantly and inversely correlated with the professional efficacy dimension of burnout and all facets
of job satisfaction.

Work-family conflict
Work interference with family life was positively correlated with family interference with work,
suggesting that although these dimensions of work-family conflict are independently related to other
variables, they are also correlated with one another. Employees sense that work interferes with family
life was also significantly positively correlated with the emotional exhaustion dimension of burnout.

Employees sense that family interferes with their work was inversely correlated with professional
efficacy, indicating that employees who feel their family has a negative impact on their work feel a
diminished sense of personal accomplishment and effectiveness. The results presented in the previous
section (Differences by parental status) suggest that this experience is likely to be significantly greater
among parents of dependent children than among other employees.

Well-being (burnout)
The emotional exhaustion dimension of burnout was significantly positively correlated with the
cynicism dimension of burnout. Cynicism was also significantly inversely correlated with employees
sense of professional efficacy.

Summary of bi-variate correlation results

These results suggest some meaningful linkages between work variables, work-family conflict, work-
related attitudes and well-being. Various aspects of job demands, including work hours during the week
and at weekends, work responsibility (for things and people) and subjective qualitative workload were
positively associated with work-to-family conflict and emotional exhaustion. Interestingly, weekday
work hours, subjective qualitative workload and both responsibility variables were positively correlated
with emotional exhaustion and professional efficacy, suggesting that although these workload variables

85
may deplete employees emotional resources, they also contribute to a positive sense of
accomplishment and effectiveness at work.

Work-family conflict was also positively related to emotional exhaustion and turnover intention,
suggesting that when employees perceive that their work interferes with family life they are more likely
to become burnt out and be inclined to quit their jobs. These results suggest that attempts to reduce
employees perceptions of work interference with family could yield positive benefits for employees and
organisations. The inverse relationship between perceived organisational support (POS) and work
interference with family, emotional exhaustion and turnover intention suggest that creating a supportive
work environment may also have positive benefits for employees and organisations.

Partnered employees
Table 44 shows the bi-variate correlations between the major family-related variables measured in the
study. The file was split and the results for partnered employees only are reported. These results are
presented once again under a series of sub-headings but readers should note that significant linkages are
only described once.

Work hours
Longer work hours were significantly correlated with difficulty in managing care of children; difficulty
in obtaining and managing family leisure time; and work-to-family conflict. Not surprisingly, higher
work hours were also associated with lower perceived work-life control. Given that working longer
hours means that time commitments interfere with family life, it is not surprising that this factor was
associated with reports of increased difficulty in managing family life, and consequently higher work-
family conflict.

Work-to-family conflict
Work-to-family conflict was positively and significantly correlated with work hours; relationship
tension; family management (care of children); family management (leisure time); and family
management (work flexibility). People who reported higher levels of work interference in family life
(whether it be strain from work, or physical work hours), also reported increased relationship tension,
and consequently poorer relationships with their partners (both sociability and companionship). In
addition, they reported lower perceived work-life and work-family control.

Work-life control
General work-life control was significantly and positively correlated with relationship sociability and
companionship with partner; and negatively correlated with relationship tension. Lower perceived
work-life control was significantly associated with a poorer spousal relationship (increased tension, less
partnership sociability; and poorer perception of partner companionship). Perceived low work-life
control was also significantly associated with poorer family management work flexibility and increased
difficulty in arranging care of children; and family leisure time. Work-family control was negatively
associated with relationship tension. People who perceived that they had low control over the impact of
work on family, had higher levels of relationship tension.

Age
As can be seen in Table 44, there were significant negative correlations between age and work hours;
age and partners work hours; and age and relationship tension. Those who were older tended to work
fewer hours, and reported their partners did likewise. Possibly as a consequence, those who were older
reported lower relationship tension.

86
Table 44 Bi-variate correlations for partnered employees

1 2 3 4 5 6 7 8 9 10 11 12 13
1. Age 1

2. Work Hours -.177(*) 1

3. Partner work hours -.282(**) -.048 1

4. Relationship sociability .085 -.132 .123 1

5. Companionship with partner .056 -.113 .064 .715(**) 1

6. Relationship tension -.248(**) .067 -.071 -.132 -.283(**) 1

7. Social undermining -.014 -.103 .082 -.140 -.399(**) .444(**) 1

8. Work-family conflict -.078 .498(**) -.156 -.304(**) -.247(**) .281(**) .031 1

9. Family-work conflict -.038 -.174(*) .015 -.072 -.099 .279(**) .262(**) .105 1

10. Family management -.330(**) .460(**) -.174 -.218(*) -.214(*) .257(**) .026 .621(**) -.036 1
care of children
11. Family management -.022 .497(**) -.077 -.300(**) -.289(**) .245(**) .021 .666(**) .042 .560(**) 1
leisure time
12. Family management .075 .374(**) -.215(*) -.123 -.215(**) .249(**) .095 .625(**) -.011 .551(**) .596(**) 1
work flexibility
13. General work-life control .091 -.510(**) .121 .286(**) .331(**) -.287(**) -.100 -.678(**) .041 -.598(**) -.697(**) -.656(**) 1

14. Work-family control .139 -.095 .167 .127 .158 -.218(*) .030 -.258(*) -.014 -.320(**) -.217(*) -.183 .263(*)
** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).

87
Family management
Poorer work flexibility, difficulty in managing care of children within the family, and difficulty in
obtaining and organising family leisure time; and were significantly associated with higher relationship
tension. In addition, they were also associated with poorer relationship sociability and partner
companionship.

Employees who are parents of dependent children

Table 45 shows the bi-variate correlations between the specific family-related variables measured in the
study. The file was split and the results for employees who were parents of dependent children are
reported. These results are presented once again under a series of sub-headings, but readers should note
that significant linkages are only described once.

Work hours
Higher work hours were significantly correlated with difficulty in managing care of children; difficulty
in obtaining family leisure time; and low work flexibility. Consequently, high work hours were
significantly associated with increased work-family conflict.

Number of children
Number of children was significantly and positively positively correlated with verbal anger towards
children. Higher number of children was also significantly associated with lower perceived support of
family. Not surprisingly, higher verbal anger was significantly associated with reduced nurturing and
support of children.

Work-to-family conflict
Work to family conflict was positively correlated with difficulty in managing care of children; difficulty
in obtaining and managing family leisure time; and difficulty in obtaining work flexibility. In addition,
high work-family conflict was associated with lower work-life control, and lower work-family control.

Work-life control
Work-life control was significantly and positively correlated with all aspects of family management.
Those reporting difficulty in managing various aspects of family life, such as care of children, leisure
time and work flexibility, also reported lower levels of work-life control.

Those who reported greater levels of work-family control also reported higher levels of support
received for the care of their child (support for family).

88
Table 45 Bi-variate correlations for parents of dependent children

1 2 3 4 5 6 7 8 9 10 11
1. Number of children 1

2. Work hours -.016 1

3. Work-family conflict -.136 .488(**) 1

4. Family-work conflict -.036 -.198(*) .079 1

5. Nurture and support child -.062 -.040 -.150 .038 1

6. Verbal anger towards child .270(**) .018 .064 .067 -.228(*) 1

7. Support for family -.209(*) .098 .062 -.114 .194 -.183 1

8. Family management
-.144 .480(**) .619(**) -.065 -.080 .018 -.053 1
care of children
9. Family management
-.072 .474(**) .650(**) .000 -.162 .014 -.001 .599(**) 1
leisure time
10. Family management
.044 .460(**) .649(**) -.059 -.116 .062 .031 .574(**) .686(**) 1
work flexibility
11. General work-life control .102 -.522(**) -.723(**) .094 .155 -.026 -.024 -.624(**) -.674(**) -.707(**) 1

12. Work-family control .022 -.110 -.250(*) -.019 .060 -.047 .255(*) -.313(**) -.217(*) -.170 .251(*)

** Correlation is significant at the 0.01 level (2-tailed).

* Correlation is significant at the 0.05 level (2-tailed).

89
Burnout analysis

Our results indicate that employees in the public and private sector construction industry organizations
surveyed experience high levels of burnout, relative to international norms. The mean burnout scores
for the current sample (Construction) are provided in Table 46 below, along with a sample of mean
burnout scores from a range of different occupations (Maslach et al., 1996). The data provided for
comparison reveals that respondents in our sample experienced higher levels of emotional exhaustion
than respondents in the military, technology and management fields. The only occupational group
indicating a higher level of emotional exhaustion than our respondents was nursing. Our respondents
also reported higher levels of cynicism and lower levels of professional efficacy than respondents in all
other occupational groups. These figures are alarming and, given that no statistically significant
differences were found between two of the three dimensions of burnout among public and private
sector respondents, it seems that burnout is a problem common to both sectors.

Table 46 Cross occupational comparison of mean burnout scores

Construction Military Technologist Management Nursing
Mean (SD) Mean (SD) Mean(SD) Mean(SD) Mean(SD)
EE 2.76 (SD=1.5) 2.05 (SD=1.23) 2.65 (SD=1.31) 2.55(SD=1.40) 2.98(SD=1.38)
CY 2.25 (SD=1.5) 1.63 (SD=1.35) 1.72 (SD=1.14) 1.32(SD=1.06) 1.80(SD=1.24)
PE 4.29 (SD=1.0) 4.60 (SD=0.93) 4.54 (SD=1.03) 4.73(SD=0.88) 4.41(SD=0.99)

Figure 52 Mean scores for the three dimensions of burnout according to job
classification

5.00 4.83
4.74
4.59
4.44
4.23 4.29
3.97 4.07
4.00

3.25
3.07
3.00 2.83 Emotional Exhaustion
Mean

2.69 2.62
2.59 2.57 2.60
2.29
2.42 2.33
Cynicism
2.16 2.22
2.09 Professional Efficacy
2.00
1.51

1.00

0.00
ot
Si

Su

C
Pr

Fo

En
on

or

he
te

oj

re

pp

gi

po
/P

ec

tra

r
ne
m

or
ro

ra
t/C

an
ct

er
tS
je

te
/A

/S

in
on
ct

er
dm

M
g
up
st

vi
En

an
Se
ru

ce
in

er
gi

ag
rv
is
ct

vi

s
ne

tra
io

so

ic

em
es
n
er

r
tio

en
M

n
an

t
ag
er

Position description

Visual inspection of Figure 52 indicates that both site engineers and employees engaged in corporate
management score more highly on the burnout dimension of emotional exhaustion. Also employees
engaged in support services and those who classify their jobs as other report higher levels of the
cynicism dimension of burnout than other employees. Foremen report much lower levels of cynicism
than employees engaged in other jobs. Figure 52 also reveals that foremen/supervisors and contract
administrators and employees engaged in other roles report the highest levels of professional efficacy.

90
Employees engaged in support services and/or corporate management along with project/construction
managers report lower levels of professional efficacy.

Work-family conflict as a mediator of the job stress-burnout relationship

As described in the introduction to this report, recent research has explored the extent to which work-
family conflict mediates the relationship between job stress variables and employee burnout, i.e. work-
family conflict is the pathway through which job stress causes employees to be burnt out. This
relationship is depicted in Figure 53.

Figure 53 Work-to-family conflict as a mediator of the job stress-burnout relationship

We tested the extent to which work-to-family conflict mediated the relationship between job stress
variables and burnout in our sample. We tested this relationship for all of the job stress variables found
to positively predict the emotional exhaustion dimension of burnout. These were: work hours;
subjective quantitative workload; and responsibility for things and people. The results are presented in
the tables below.

Table 47 shows the results of statistical testing procedures to determine whether work-to-family
conflict mediates the relationship between work hours and emotional exhaustion. The results show that
the conditions for partial mediation are met, i.e:

Work hours significantly affects work-to-family conflict in the first equation;

Work hours significantly affects emotional exhaustion in the second equation;

Work-to-family conflict significantly affects emotional exhaustion in the third equation; and

The effect of work hours on emotional exhaustion in the third equation is less than in the
second equation.

Thus, work-to-family conflict partially mediates the relationship between work hours and emotional
exhaustion.

91
Table 47 Regression analysis examining work-to-family conflict as a mediator in the
work hours-emotional exhaustion relationship
Step Variable B SE p
Equation 1: Work-to-family Constant 1.436 .574 .013
conflict regressed on average Average hours worked .061 .012 .343 .000
hours worked
Equation 2: Emotional Constant .985 .479 .041
exhaustion regressed on Average hours worked .038 .010 .262 .000
average hours worked
Equation 3: Emotional Constant .669 .470 .157
exhaustion regressed on Average hours worked .024 .010 .169 .017
average hours worked and
work-to-family conflict Work-to-family conflict .221 .057 .272 .000

Table 48 shows the results of statistical testing procedures to determine whether work-to-family
conflict mediates the relationship between subjective quantitative workload and emotional exhaustion.
The results show that the conditions for partial mediation are met, i.e:

Subjective quantitative workload significantly affects work-to-family conflict in the first

equation;

Subjective quantitative workload significantly affects emotional exhaustion in the second
equation;

Work-to-family conflict significantly affects emotional exhaustion in the third equation; and

The effect of subjective quantitative workload on emotional exhaustion in the third equation is
less than in the second equation (as indicated by a lower coefficient).

Thus, work-to-family conflict partially mediates the relationship between subjective quantitative
workload and emotional exhaustion.

Table 48 Regression analysis examining work-to-family conflict as a mediator in the

subjective quantitative workloademotional exhaustion relationship
Step Variable B SE p
Equation 1: Work-to-family Constant .353 .977 .718
conflict regressed on Subjective quantitative 1.142 .278 .279 .000
subjective quantitative workload
workload
Equation 2: Emotional Constant -2.012 .752 .008
exhaustion regressed on Subjective quantitative 1.376 .241 .414 .000
subjective quantitative workload
workload
Equation 3: Emotional Constant -2.079 .732 .005
exhaustion regressed on Subjective quantitative 1.161 .217 .349 .000
subjective quantitative workload
workload and work-to-family Work-to-family conflict .189 .053 .232 .000
conflict

Table 49 shows the results of statistical testing procedures to determine whether work-to-family
conflict mediates the relationship between responsibility (for things) and emotional exhaustion. The
results show that the conditions for full mediation are met, i.e:

Responsibility (for things) significantly affects work-to-family conflict in the first equation;

Responsibility (for things) significantly affects emotional exhaustion in the second equation;

Work-to-family conflict significantly affects emotional exhaustion in the third equation; and

The effect of responsibility (for things) on emotional exhaustion in the third equation is no
longer significant.

92
Thus, work-to-family conflict fully mediates the relationship between responsibility (for things) and
emotional exhaustion.

Table 49 Regression analysis examining work-to-family conflict as a mediator in the

responsibility (for things)-emotional exhaustion relationship

Step Variable B SE p
Equation 1: Work-to-family Constant 1.498 .613 .015
conflict regressed on Responsibility (for things) .823 .174 .317 .000
responsibility (for things)
Equation 2: Emotional Constant 1.618 .518 .002
exhaustion regressed on Responsibility (for things) .338 .147 .160 .023
responsibility (for things)
Equation 3: Emotional Constant 1.241 .503 .015
exhaustion regressed on Responsibility (for things) .131 .149 .062 .381
responsibility (for things) and
work-to-family conflict Work-to-family conflict .252 .057 .310 .000

Table 50 shows the results of statistical testing procedures to determine whether work-to-family
conflict mediates the relationship between responsibility (for persons) and emotional exhaustion. The
results show that the conditions for partial mediation are met, i.e:

Responsibility (for persons) significantly affects work-to-family conflict in the first equation;

Responsibility (for persons) significantly affects emotional exhaustion in the second equation;

Work-to-family conflict significantly affects emotional exhaustion in the third equation; and

The effect of responsibility (for persons) on emotional exhaustion in the third equation is less
than in the second equation.

Thus, work-to-family conflict partially mediates the relationship between responsibility (for persons)
and emotional exhaustion.

Table 50 Regression analysis examining work-to-family conflict as a mediator in the

responsibility (for persons)-emotional exhaustion relationship
Step Variable B SE p
Equation 1: Work-to-family Constant 2.648 .436 .000
conflict regressed on Responsibility (for .542 .122 .301 .000
responsibility (for persons) persons)
Equation 2: Emotional Constant 1.513 .359 .000
exhaustion regressed on Responsibility (for .369 .100 .252 .000
responsibility (for persons) persons)
Equation 3: Emotional Constant .954 .373 .011
exhaustion regressed on Responsibility (for .246 .101 .168 .016
responsibility (for persons) persons)
and work-to-family conflict
Work-to-family conflict .227 .056 .279 .000

The implication of these findings for the prevention and mitigation of burnout are discussed in Section 4
of this report.

93
Job control/flexibility

A series of multiple regression analyses was performed in order to identify potential moderator
variables. Moderator variables are variables that affect the direction and/or strength of the relationship
between an independent/predictor variable and a dependent variable (Baron & Kenny, 1986). See also
Data analysis in the Section 2 of this report.

Regression analyses were performed to explore the relationship between work hours, job
flexibility/control and employees perceived work-to-family conflict. These analyses explored both the
main effects and moderating effects of job flexibility/control on work-to-family conflict. In previous
research, the extent to which employees have flexibility and control over their work arrangements has
been linked to the extent to which their work negatively interferes with their family life. Batt and
Valcour (2003) note that professional and managerial employees are likely to have greater demands on
their time. However, professionals and managers are also likely to have greater opportunities to exert
control in managing their time and schedules. It is likely that work-to-family conflict will be inversely
related to job flexibility/control owing to the fact that increased job flexibility/control will enable
employees to manage their work commitments so as to minimise disruption to non-work activities.
Thus our first hypothesis relating to job flexibility/control is as follows:
H1: There will be a negative relationship between job flexibility/control and work-to-family
conflict.
It was also considered possible that job flexibility/control could have a buffering effect, such that the
relationship between work schedule demands and work-to-family conflict will be dependent upon the
extent to which an individual enjoys job flexibility/control. Thus, where control/flexibility is high, the
relationship between work hours and work-family conflict will be weaker than where job
flexibility/control is low. Specifically it was hypothesised that the relationship between work hours and
work-to-family conflict would be moderated by the extent to which employees perceive that they have
control or flexibility in determining their work arrangements. This hypothesis is also consistent with
Karaseks Demand-Control theory of stress, which holds that excessive job demands have a more
damaging effect in conditions of low employee control. Thus, hypotheses 2 and 3 were as follows:
H2: The relationship between work hours and work-to-family conflict will be weaker where
employees perceive that they have flexibility/control in their work arrangements.

This moderation model is depicted in Figure 54.

Figure 54 Job control/flexibility as a moderator of the relationship between job

demands and work-to-family conflict

The results of these regression analyses are shown in Table 51.

94
Table 51 Multiple regression for work hours and work-life control as predictors of
work-to-family conflict
2
Step and predictors R F-change dfs
1. Controls .041 8.618** 200
a
Industry sector -.203**
2. Independent variables .183 23.325*** 198
Average weekly work hours (WH) .185*
Work-life control (general) (WLC) -.387***
3. Interaction .000 .004 197
WH x WLC .004
a 1= private sector, 2= public sector
* p=< .05, ** p =< .01, *** p=<.001

Table 51 shows the results of the regression analysis relating to average weekly work hours and
employees general control over work-life issues. This general control tapped items, such as the extent
to which employees could choose to do some of their work at home or choose when to begin and end
each work week. Prior to entering the independent variables or the interaction term, industry sector was
entered as a control variable. This was entered as a control because industry sector was significantly
correlated with work-to-family conflict. Table X shows that both average weekly work hours and
general work-life control significantly predict work-to-family conflict. The higher the number of hours
worked the greater the work-to-family conflict. Also, the relationship between work-life control and
work-to-family conflict was in the expected direction, i.e., the greater the degree of work-life control,
the lower the experienced work-to-family conflict. Thus, hypothesis one was supported. However,
there was no significant interaction effect indicating that work-life control does not moderate the effect
of work hours on work-to-family conflict in our sample. Thus, hypothesis 2 was not supported by this
analysis.

Table 52 Multiple regression for work hours and work-family control as predictors of
work-to-family conflict
2
Predictors R F-change dfs
1. Controls .146 16.103 94
a
Industry sector -.382***
2. Independent variables .150 9.782 92
Average weekly work hours (WH) .428***
Work-life control (general) (WLC) -.205*
3. Interaction .006 .727 91
WH x WLC .082
a 1= private sector, 2= public sector
* p=< .05, ** p =< .01, *** p=<.001

Table 52 shows the results of the regression analysis relating to average weekly work hours and
employees control over work and family issues. These questions related only to respondents who
indicated they had parenting responsibilities. This family control variable tapped items such as the
extent to which employees have choice over the way they balance work and parenting, and over the
care they provide for children or other dependent relatives. Prior to entering the independent variables
or the interaction term, industry sector was entered as a control variable. This was entered as a control
because industry sector was significantly correlated with work-to-family conflict. Table 52 shows that
both average weekly work hours and work-family control significantly predict work-to-family conflict.
The higher the number of hours worked the greater the work-to-family conflict. Also, the relationship
between work-family control and work-to-family conflict was in the expected direction, i.e., the greater
the degree of work-family control, the lower the experienced work-to-family conflict. Thus, hypothesis
1 was supported. However, there was no significant interaction effect which indicates that work-family
control does not moderate the effect of work hours on work-to-family conflict in our sample. Thus,
hypothesis 2 was not supported by this analysis.

95
Supportive work environments, work-to-family conflict and burnout

Substantial research has examined the main and buffering effects of support in the stressor-strain
relationship. These studies have supported both main and buffering effects. (See, for example, Casper
et al 2002). On the basis of this previous research, we undertook a series of regression analyses to
determine whether different types of support in the work environment (i.e. perceived organisational
support, supervisor support and co-worker support) moderated the relationship between employees
work-to-family conflict and emotional exhaustion (burnout). These regression analyses are presented
below. Significant interactions were plotted by calculating regression lines using procedures described
by Cohen, Cohen, West and Aiken (2003). Regression lines were plotted at high (+1 SD), average (SD)
and low (-1 SD) levels of the moderator variable.

Perceived organisational support

Previous research has linked a lack of supportive in the work environment to burnout (Houkes et al
2001). It is suggested that a lack of support contributes to emotional exhaustion because opportunities
to benefit from the protective effects of positive social contacts are limited. We suggest that when
organisational support is perceived to be lacking, employees may feel insecure about their ability to
secure needed support or resources to perform their work and this triggers emotional exhaustion. Thus
our first hypothesis is as follows:

H1: There will be a negative relationship between POS and emotional exhaustion.

Previous analysis of our data revealed that work-to-family conflict mediates the relationship between
job schedule demands and the emotional exhaustion dimension of burnout (See burnout, on page X of
this report). Thus there is a direct pathway linking work-to-family conflict and emotional exhaustion.
That is, when employees experience high levels of work-family conflict they will suffer from emotional
exhaustion. However, we posit that when employees work in a supportive organisational environment,
this will reduce the stress associated with work-to-family conflict. This stress reduction could arise, for
example, because the organization permits some degree of flexibility to accommodate family demands
during work time. Thus perceived organisational support (POS) is likely to reduce the stress associated
with work-to-family conflict and mitigate or buffer the effects of work-to-family conflict on emotional
exhaustion. Thus second our hypothesis is as follows:

H2: The relationship between work-to-family conflict and emotional exhaustion will be weaker
where POS is high.
This hypothesis is depicted in Figure 55 below:

Figure 55 POS as a moderator of the relationship between work-to-family conflict and

emotional exhaustion

96
Initially we tested whether emotional exhaustion differed by industry sector, age or gender. No
significant differences existed so we did not include any of these variables in the regression analysis as
controls.

Table 53 Multiple regression for work-to-family conflict and POS as predictors of

emotional exhaustion (burnout)
2
Step and predictors R F- change dfs
1. Independent variables
Work-to-family conflict (WFC) .247 32.575 199 .262***
Perceived organizational support (POS) -.378***
2. Interaction
.020 5.515 198
WFC x POS -.147*
* p=< .05, ** p =< .01, *** p=<.001

Table 53 shows a main effect of POS on emotional exhaustion in the expected direction. Thus,
hypothesis H1 was supported. The interaction between work-to-family conflict and POS also had a
significant effect on emotional exhaustion, indicating that POS moderates the relationship between
work-to-family conflict and emotional exhaustion. The nature of this moderation effect is shown in
Figure 56 below.

Figure 56 The moderating effect of perceived organizational support on the relationship

between work to family conflict and emotional exhaustion (burnout)

4.5

4
Emotional exhaustion (burnout)

3.5

2.5

1.5

0.5

0
1 2 3 4 5 6 7
Work-to-family conflict

Low POS Medium POS High POS

Figure 56 shows that when perceived organizational support (POS) is high, the relationship between
work-to-family conflict and emotional exhaustion (burnout) is relatively weak. However, when the level
of POS is low, the slope of the regression line is steeper, representing a stronger positive relationship
between work-to-family conflict and emotional exhaustion (burnout).

Supervisor support
A series of regression analyses was also conducted to explore the relationship between supervisory
support of employees family life and emotional exhaustion. These analyses explored both the main
effects and buffering effects of supervisor support on employees burnout. Supervisors play an

97
important role in the decentralized construction industry and it was anticipated that when supervisors
recognise that employees have obligations outside of work and are supportive of these obligations,
employees will be less prone to burnout. Thus our first hypothesis was as follows:

H1: There will be a negative relationship between supervisor support and burnout.

It was also considered possible that supervisor support of employees personal lives could have a
buffering effect, such that where supervisor support is high, the relationship between work-family
conflict and burnout will be weaker than where supervisor support is low. ODriscoll et al (2003)
examined supervisor support and report than when work to family interference is high, respondents
who perceive high levels of supervisor support experience less psychological strain than those who
perceive low supervisor support. Specifically it was hypothesized that the relationship with between
work-to-family conflict and the emotional exhaustion dimension of burnout would be moderated by
the extent to which employees receive social support from their supervisor. Thus our second
hypothesis was follows:

H2: The relationship between work-to-family conflict and emotional exhaustion will be weaker
where employees receive social support from their supervisor.

This moderation model is depicted in Figure 57.

Figure 57 Supervisor support as a moderator of the relationship between work to family

conflict and emotional exhaustion

The model was tested for each of the three dimensions of supervisory support revealed in the Principal
Components Analysis (i.e. supervisors emotional supportiveness, supervisors practical supportiveness
and supervisors criticism/resentment of employees family obligations).

The results of the regression analyses are shown in Table 54, 55 and 56 below:

Table 54 Multiple regression for work-to-family conflict and supervisor emotional

support as predictors of emotional exhaustion
2
Step and predictors R F- change dfs
1. Independent variables
Work-to-family conflict (WFC) .134 15.393 199 .299***
Supervisor Support emotional (SS-E) -2.419*
2. Interaction
.008 1.799 198
WFC x SS-E -.093
* p=< .05, ** p =< .01, *** p=<.001

98
Table 54 reveals that the emotional dimension of supervisor support had a main effect on emotional
exhaustion in the expected direction, i.e, employees whose supervisors provide more emotional support
are less emotionally exhausted. However, the emotional dimension of supervisor support did not
significantly moderate the relationship between work-family conflict and emotional exhaustion. Thus,
for this dimension of supervisor support the first hypothesis was supported but the second hypothesis
was not supported.

Table 55 Multiple regression for work-to-family conflict and supervisor practical

support as predictors of emotional exhaustion
2
Step and predictors R F- change dfs
1. Independent variables
Work-to-family conflict (WFC) .120 13.612 199 .308***
Supervisor Support emotional (SS-P) -.111
2. Interaction
.073 17.947 198
WFC x SS-P -.285***
* p=< .05, ** p =< .01, *** p=<.001

Table 55 reveals that the practical dimension of supervisor support did not have a main effect on
emotional exhaustion. However, the interaction between supervisors practical support for employees
work-family balance and work-to-family conflict was highly significant, indicating that practical support
from ones supervisor moderates the relationship between work-to-family conflict and emotional
exhaustion. Thus, for this dimension of supervisor support the first hypothesis was not supported but
the second hypothesis was supported. The nature of this moderation effect is shown in Figure 58
below.

Figure 58 The moderating effect of practical support from ones supervisor on the
relationship between work to family conflict and emotional exhaustion
(burnout)

4.5

4
Emotional exhaustion (burnout)

3.5

2.5

1.5

0.5

0
1 2 3 4 5 6 7
Work-to-family conflict

Low SS-Practical Medium SS-Practical High SS-Practical

Figure 58 shows that when the level of practical support received from ones supervisor is high, the
relationship between work-to-family conflict and emotional exhaustion (burnout) is relatively weak.
However, when the level of practical supervisor support is low, the slope of the regression line is

99
steeper, representing a stronger positive relationship between work-to-family conflict and emotional
exhaustion (burnout).

Table 56 reveals that supervisors criticisms or resentment of employees family commitments were
found to have a significant main effect on emotional exhaustion. This effect was in the expected
direction, i.e. employees whose supervisors are critical/resentful of their family commitments are more
emotionally exhausted. Supervisors criticism and resentment was also found to have a moderate the
relationship between work-to-family conflict and emotional exhaustion. Thus, for this dimension of
supervisor support both the first and the second hypotheses were supported. The nature of this
moderation effect is shown in Figure 59 below.

Table 56 Multiple regression for work-to-family conflict and supervisor

criticism/resentment as predictors of emotional exhaustion
2
Step and predictors R F- change dfs
1. Independent variables
Work-to-family conflict (WFC)
.124 14.124 199 .312***
Supervisor Support criticism and resentment (SS-
.060*
CR)
2. Interaction
.042 9.971 198
WFC x SS-CR .201**
* p=< .05, ** p =< .01, *** p=<.001

Figure 59 The moderating effect of supervisor criticism/resentment of employees

family obligations on the relationship between work to family conflict and
emotional exhaustion (burnout)

4.5

4
Emotional exhaustion (burnout)

3.5

2.5

1.5

0.5

0
1 2 3 4 5 6 7
Work-to-family conflict

Low SS - Criticism Medium SS - Criticism High SS - Criticism

Figure 59 shows that when supervisors criticism/resentment of employees family obligations is low,
the relationship between work-to-family conflict and emotional exhaustion (burnout) is relatively weak.
However, when supervisors criticism/resentment is high, the slope of the regression line is much
steeper, representing a stronger positive relationship between work-to-family conflict and emotional
exhaustion (burnout).

100
Co-worker support
A series of regression analyses was also conducted to explore the relationship between the support
employees receive from their co-workers and emotional exhaustion. Again, these analyses explored
both the main effects and buffering effects of co-worker support on employees burnout. Thus our first
hypothesis was as follows:

H1: There will be a negative relationship between co-worker support and burnout.

It was also considered possible that co-worker support of could have a buffering effect, such that
where co-worker support is high, the relationship between work-family conflict and burnout will be
weaker than where co-worker support is low. Specifically it was hypothesized that the relationship with
between work-to-family conflict and the emotional exhaustion dimension of burnout would be
moderated by the extent to which employees receive social support from their co-workers. Thus our
second hypothesis was follows:

H2: The relationship between work-to-family conflict and emotional exhaustion will be weaker
where employees receive social support from their co-workers.

This hypothesis is depicted in Figure 60 below:

Figure 60 Co-worker support as a moderator of the relationship between work-to-family

conflict and emotional exhaustion

The model was tested for each of the two dimensions of co-worker support revealed in the Principal
Components Analysis (i.e. emotional support and practical support).

The results of the regression analyses are shown in Tables 57 and 58 below:

Table 57 Multiple regression for work-to-family conflict and co-worker emotional

support as predictors of emotional exhaustion
2
Step and predictors R F- change dfs
1. Independent variables
Work-to-family conflict (WFC) .123 13.976 199 .333*
Co-worker Support emotional (CWS-E) -.121*
2. Interaction
.007 1.539 198
WFC x CWS-E -.083
* p=< .10, ** p =< .01, *** p=<.001

Table 57 reveals that emotional support from co-workers was found to have a significant main effect
on emotional exhaustion. This effect was in the expected direction, i.e. employees whose co-workers

101
are supportive are less emotionally exhausted. However, co-workers emotional support was not found
to moderate the relationship between work-to-family conflict and emotional exhaustion. Thus for
emotional support from co-workers, the first hypothesis was supported but the second hypothesis was
not.

Table 58 Multiple regression for work-to-family conflict and co-worker practical

support as predictors of emotional exhaustion
2
Step and predictors R F- change dfs
1. Independent variables
Work-to-family conflict (WFC) .116 13.025 199 .327***
Co-worker Support practical (CWS-P) -.085
2. Interaction
.047 11.132 198
WFC x CWS-P -.238***
* p=< .05, ** p =< .01, *** p=<.001

Table 58 reveals that practical support from co-workers, for example in being trusted to lend a hand
with work tasks if necessary, was found to have no significant main effect on emotional exhaustion.
However, co-workers practical support did moderate the relationship between work-to-family conflict
and emotional exhaustion. Thus for practical support from co-workers, the first hypothesis was not
supported but the second hypothesis, i.e. the moderation effect, was supported. The nature of this
moderation effect is shown in Figure 61 below.

Figure 61 The moderating effect of co-worker practical support on the relationship

between work to family conflict and emotional exhaustion (burnout)

4.5

4
Emotional exhaustion (burnout)

3.5

2.5

1.5

0.5

0
1 2 3 4 5 6 7
Work-to-family conflict

Low CWS - Practical Medium CWS - Practical High CWS - Practical

Figure 61 shows that when co-workers provide practical support to employees, the relationship
between work-to-family conflict and emotional exhaustion (burnout) is relatively weak. However, when
the level of practical support from co-workers is low, the slope of the regression line is steeper,
representing a stronger positive relationship between work-to-family conflict and emotional exhaustion
(burnout).

102
The nature, antecedents and outcomes of work-family conflict

Cross-role effects
Contemporary models of the work-family interface that take a bi-directional approach which recognises
that work can impact upon family life and also that family life can impact upon work. Recent research
demonstrates that the two types of work-family conflict are associated with different work and family
related antecedents and outcomes.

As described in the introduction of this report, models of the work-family interface have used work-
family conflict as a key mediating variable accounting for cross-role relationships between work and
family life (Frone et al., 1997). Thus, work is believed to affect family life and family life is believed to
affect work through the work-family conflict experienced by an individual. Perhaps the most oft-cited
model of the work-family interface is that developed by Frone et al. Frone et al.s model uses a bi-
directional approach to work-family conflict and also proposes domain-specific antecedents and
outcomes for the two types of conflict. The basic components of this model are presented briefly
below.

Frone et al. (1997) draw a distinction between proximal (i.e. direct) and distal (i.e. indirect) predictors of
work-family conflict, suggesting that proximal predictors mediate the relationship between distal
predictors and conflict. In our analysis we considered only proximal predictors in the first instance to
determine whether Frone et al.s propositions about domain-specific antecedents and outcomes of
work-family conflict were supported in our sample.

Frone et al. (1997) identified three types of proximal predictor of work-family conflict. Time-based
predictors represent time commitments to one role that limit an individuals ability to participate in
another role. Strain-based predictors are role-related dissatisfaction or distress. It is argued that time
commitments to work, or role-related dissatisfaction in the work domain, predict work-to-family
conflict, whereas family-time commitments or family-related distress lead to family-to-work conflict.
Frone et al. (1997) also identify role overload as a predictor of work-family conflict. Again, they posit
domain specific effects suggesting that work overload will predict work-to-family conflict and family
overload will lead to family-to-work conflict.

Frone et al. (1997) also suggested that the outcomes of both types of work-family conflict would be
domain-specific. They suggested that work-to-family conflict would be a direct and positive predictor of
family distress and family-to-work conflict would be a direct and positive predictor of work distress.
They also suggest that work-family conflict would predict domain-specific performance, such that
work-to-family conflict would be a direct and negative predictor of performance in the family domain
and family-to-work conflict would be a direct and negative predictor of performance in the work
domain. The basis for this suggestion is that the quality of life associated with one role and the ability to
perform in that role will be undermined by an inability to participate fully in that role because of
interference from another role.

This simplification of Frone et al.s model is depicted in the Figure 62.

103
Figure 62 Conceptual model of the work-family interface (adapted from Frone et al.
1997)

In order to test this model (or theory) we used a statistical technique known as path analysis. Path
analysis uses multiple regression equations to calculate the relationships between the model variables.
The magnitude of the pathways is then examined to determine how variables in the model impact on
each other and the outcome variables (in the above case outcome variables are work role performance
and work distress/dissatisfaction and family role performance and family distress/dissatisfaction).

In the model, work-role performance was conceptualised as an individuals stated turnover intention.
Frone et al. (1997) also used turnover intention as a dimension of work-role behaviour. We
conceptualised work distress as the emotional exhaustion dimension of burnout, which is regarded as a
strain outcome arising from exposure to stressors in the work environment. Family-role performance
was conceptualized as the quality of an individuals relationship with his/her spouse/partner. Family
distress was conceptualised as conflict experienced in the relationship with ones spouse/partner.
Owing to the fact that we did not measure some of the family variables directly, we used proxy
measures for some elements of the model. Thus, the number of children present was used as a proxy
for family workload and the number of work days taken during the preceding twelve months to care for
a dependant was used as a proxy measure of time commitment to family.

The results of the model testing are shown in Figure 63.

104
Figure 63 Summary of standardised path co-efficients for work-family conflict model

All hypothesised work-domain predictors of work-to-family conflict were highly significant (p = <.001)
and in the expected direction. Thus, emotional exhaustion (burnout), subjective overload and average
hours worked per week all positively predict employees work-to-family conflict in our sample.
However, only one of the hypothesised family-domain predictors of family-to-work conflict was
significant. This was conflict in relationship with ones spouse or partner (p = .001). This finding is
inconsistent with the results of Frone et al. (1997) who report that both family distress and family
overload were both significant predictors of family-to-work conflict. The lack of significant findings
could be because we used proxy measures of family time commitment and family overload, and
therefore it should not be assumed that family-domain predictors would not have a stronger predictive
ability for family-to-work conflict were more direct measures used. Both of the hypothesised family-
domain outcomes of work-to-family conflict were highly significant (p = < .001) and in the expected
direction. However, neither of the hypothesised work-domain outcomes of family-to-work conflict was
significant.

The magnitude of these pathways can be determined by multiplying pathway coefficients together. This
process is known as tracing (Cohen et al., 2003). Thus, the indirect effect of hours worked per week on
tension in relationship with spouse/partner was: .477 x .264 = .126 (the direction of this pathway is not
as important here as the magnitude). The indirect effect of subjective qualitative workload on
relationship tension was: .440 x .264 = .11;, and the indirect effect of emotional exhaustion on
relationship tension was .454 x .264 = .120. The indirect effect of average hours worked each week on
the quality of the relationship with ones spouse/partner was: .477 x -.305 = -.145 (again, the direction
of this pathway is not as important as the magnitude). The indirect effect of subjective qualitative
workload on relationship quality was .440 x -.305 = -.134. The indirect effect of emotional exhaustion
on relationship quality was .454 x -.305 = -.138.

105
Thus our results lend partial support to the theory that work-family conflict is a key mediating variable
accounting for cross-role relationships between work and family. The results also lend some support to
the domain-specific antecedents and outcomes of work-family conflict. However, in our sample, it
appears that these cross-role relationships are asymmetrical. In our sample it seems that family life is
highly susceptible to negative interference from work (via work-to-family conflict), but that work
domain outcomes are not significantly impacted by family interference with work.

Within-role effects
Owing to the fact that the first model received only partial support, i.e. the hypothesized relationship
between family-to-work conflict, turnover intention and work distress (emotional exhaustion) were
non-significant, we decided to test an alternative model (or theory) relating to the outcomes of work-
family conflict. In this alternative theory we hypothesised that work-to-family conflict would be
predicted by work hours, subjective qualitative overload and burnout (as in the first model). However,
we hypothesised that work-to-family conflict would predict work-related outcomes, including
organisational commitment and turnover intention. Allen et al. (2000) note that work-to-family conflict
is associated with work-related, non-work-related and stress-related undesirable outcomes.

In particular, we hypothesised that work-to-family conflict would positively predict turnover attention
because when employees experience work-to-family conflict they are likely to withdraw from work in
an attempt to eliminate the conflict. Boyar et al. (2003) suggest that when work interferes with
competing family obligations, causing work-to-family conflict, work is seen as the proximal cause of
employees thoughts about quitting in order to eliminate this conflict.

In addition, because the source of work-to-family conflict is the work domain, it would be reasonable
to assume that high levels of work-to-family conflict would have less positive feelings about their
organisation and be less emotionally attached to their organisation. Thus we hypothesise that work-to-
family conflict will be inversely related to affective organisational commitment and intrinsic job
satisfaction, i.e. satisfaction with the respect received from people at the workplace. In the pilot study
of employees of another large contracting organisation, Baulderstone Hornibrook, Lingard and Francis
(2002) also found that perceived fairness of the exchange relationship was related to work hours and
family issues. Thus, employees who worked long hours and who consequently experienced work
interference with family life were often unhappy about the reward they received relative to this input to
the organisation. Thus we hypothesised that work-to-family conflict would be inversely related to
extrinsic job satisfaction, i.e. the satisfaction with the amount of pay and benefits an individual receives.

Casper et al. (2002) draw on cognitive dissonance theory, which holds that if an individual views
inconsistencies between their beliefs and behaviours, they experience an uncomfortable psychological
state until they change their belief or behaviour to create consistency. They suggest that employees who
experience high levels of work-to-family conflict but remain with the organisation experience conflict
and guilt for keeping a job that disrupts their family life. However, these individuals reconcile this
uncomfortable state by attributing their reason for staying in their jobs as needs-based, i.e. because the
have to, rather than because they want to. Thus, we hypothesise that work-to-family conflict will be
positively related to the continuance dimension of organisational commitment.

The alternative model specifying within-domain antecedents and outcomes of work-to-family conflict is
shown in the Figure 64.

106
Figure 64 Alternative model of within-domain antecedents and outcomes of work-to-
family conflict showing standardised path coefficients

Intrinsic
job satisfaction

Emotional NS
Exhaustion
(burnout) .454 Extrinsic job
-.154 satisfaction

Subjective .440 Work-to-family .147

Intention to
quantitative conflict turnover
workload
NS

Affective
Average .477 commitment
hours worked -.158
per week

Continuance
commitment

Figure 64 shows the standardised path coefficients for this model. All hypothesised work-domain
predictors of work-to-family conflict were highly significant (p=<.001) and in the expected direction.
Thus, emotional exhaustion (burnout), subjective overload and average hours worked per week all
positively predict employees work-to-family conflict in our sample. However, contrary to our
hypothesis, work-to-family conflict did not significantly predict employees affective commitment or
intrinsic job satisfaction in our sample. Work-to-family conflict was a significant predictor of
continuance commitment but the direction of the relationship was not as expected. Thus, employees
who experienced higher levels of work-to-family conflict experienced lower, rather than higher, levels
of continuance commitment.

Only two work-related outcome variables in the model were significantly predicted by work-to-family
conflict in the directions hypothesised. These were turnover intention and extrinsic job satisfaction.
Employees who experienced higher levels of work-to-family conflict expressed lower levels of extrinsic
job satisfaction and a stronger intention to quit their jobs than employees who experienced lower levels
of work-to-family conflict.

107
Results of Partners Questionnaire

Introduction
A total of 35 respondents answered the on-line survey. Twelve respondents chose to download the
PDF version of the survey, available on the website, and returned a completed hard copy to the
research team. All 47 completed questionnaires were considered, representing six (12.7%) respondents
from the private sector organisation and 41 (87.2%) respondents from the public sector organisation.
Assuming half of the employee sample were partnered this represents a response rate of approximately
11% from the private sector organisation and 16% from the public sector organisation. The response
rate was disappointing and, as such, only some descriptive statistics and bi-variate correlations have
been used and the results should be viewed cautiously.

Demographic Information
Before presenting the results of the data analysis, the characteristics of the sample are described. Table
59 showing the demographic characteristics of the sample. These are then discussed in more detail
below.

Table 59 Demographic characteristics of sample

N % N %
Age Employment
20 29 years 9 19.1 In paid employment 29 61.7
30 39 years 14 29.8 Not in paid employment 18 38.3
40 49 years 8 17.0
50 59 years 13 27.7 Hours worked per week
60 year + 3 6.4 Up to 9 hours 2 6.9
10 to 19 hours 6 20.7
Gender 20 to 29 hours 4 13.8
Male 5 10.6 30 to 39 hours 11 37.9
Female 42 89.4 40 to 49 hours 6 20.7
50+ hours 0 0
Dependent children
Yes 26 56.5 Work role
No 20 43.5 Manager/administrator 3 6.3
Professional 14 29.2
Partners employer Associate professional 2 4.2
Private 6 12.8 Clerical or intermediate service worker 11 22.9
Public 41 87.2

Age and gender of the sample

Respondents were asked to indicate their year of birth. From this, the mean age of the sample was year
calculated to be 41.85 years (SD= 11.26) with a range of 20 to 65 years. The composition of the sample
(age cohorts) is illustrated in the Figure 65.

108
Figure 65 Age cohorts of all respondents

6.38%
60 years + 19.15%
20 to 29 years

27.66%
50 to 59 years

29.79%
30 to 39 years

17.02%
40 to 49 years

The mean age of the respondents was viewed alongside their partners sector of employment. The
mean ages of the two groups were found to be quite different, with the public sector having an older
sample (43.39 years, SD= 11.07), compared to 31.33 (SD= 5.85) for private sector. Given the low
sample size, this information should be viewed conservatively.

Figure 66 provides a breakdown of partners age cohorts by employment sector.

Figure 66 Respondents age by partners sector of employment

14
13 13

12
Company
10 Private sector
Public sector
Count

8
7

6
5
4
4
3

2
1 1

0
20 to 29 years 30 to 39 years 40 to 49 years 50 to 59 years 60 years +
Age Cohort

Of all respondents five (10.6%) were male and 42 (89.4%) were female. The figure below shows the
percentage of males and femaless by the partners employment sector.

109
Figure 67 Percentage of male and female respondents by partners employment sector

100.0%
90.2%
83.3%
80.0%

Company
Percent

60.0%
Private sector
Public sector
40.0%

20.0% 16.7%
9.8%

0.0%
male female
Gender

The average age of the male respondents was 45.8 years (SD=17.71) and female respondents was 41.38
years (SD=10.46). See Figure 68.

Figure 68 Male and female respondents by age cohorts

14

12

10

Gender
Count

8
male
female
6

0
20 to 29 years 30 to 39 years 40 to 49 years 50 to 59 years 60 years +
Age Cohorts

Employment

Participation in paid workforce

Participants were asked to indicate whether they worked in paid employment. Eighteen (38.3%) stated
they did not, while 29 (61.7%) reported working in paid employment. Of those who indicated they
worked, three (10%) were employed in a managerial/administrative position; 14 (46.7%) in a
professional capacity; two (6.7%) as associate professionals; and11 (36.7%) in a clerical or intermediate
service worker position (please refer to Table 60).

110
Table 60 Work role, gender and partners employment sector

Company Gender
Total
sample Private Public
Male Female
sector sector
Manager/administrator 3 2 1 1 2
Professional 14 2 12 2 12
Associate professional 2 0 2 0 2
Clerical or intermediate service worker 11 0 11 0 11

While the average number of hours of paid employment worked by the respondent was 28.59 (SD =
12.51), the range was very large (4 hours 45 hours) (See Figure 69).

Figure 69 Hours worked in paid employment per week

6.9%
20.69% up to 9 hours
40 to 49 hours

20.69%
10 to 19 hours

13.79%
20 to 29 hours

37.93%
30 to 39 hours

A further breakdown for gender was carried out. The average hours worked per week for male
respondents was 35.67 (SD=4.93), and female respondents 27.77 (SD=12.9). Again this information
should be viewed cautiously due to low male participation rates.

Time off work

Figure 70 outlines the number of days those in paid employment took off as holidays in the past 12
months.

111
Figure 70 Days off for holidays

8
8

5 5
Count

4
4

3 3

0
no days 1 - 10 days 11 - 20 21 - 30 41 - 50 51 - 60 61 or more
off for off for days off days off days off days off days off
holidays holidays for for for for for
holidays holidays holidays holidays holidays
Holiday time grouping

Respondents were asked how often they had seven or more consecutive days off. The mean value was
2.55 times per year (SD=2.789) ranging from 0 to 15 times per year.

The mean number of days taken off to care for sick dependants was 14.48 days in the past 12 months
(SD=46.705). This ranged from 0 to 248 days with almost half the sample (48.3%) taking no days off.
While the male sample was very small it is interesting to note that the mean number of days to take care
of sick dependants was 3.33 days for males and 15.33 for females.

Work involvement
Work involvement was determined by asking respondents to indicate the extent to which particular
statements reflects their involvement with their work. A 5-point Likert scale was used ranging from
strongly agree (1) to strongly disagree (5).

As items were obtained originally from two separate measures, a factor analysis was undertaken to
ensure the new scale was measuring the same construct. All items loaded on one factor except for the
fifth item My job is a very small part of myself (this was reversed scored and the new reversed value
was used in the factor analysis). The Cronbachs alpha coefficient for the whole scale was .89 and .91
when the item was excluded. While both values indicated an acceptable level of internal consistency
reliability, it was decided to remove this fifth item from the scale.

The mean values for each item are shown below. A higher score reflects an individuals view that their
work was very important to them. The average score for the new scale, excluding the fifth item, was 2.71
(SD=.743) with an alpha coefficient of .91.

112
Table 61 Work involvement
Mean Std.
Value Deviation
If I were to rank (in importance to me) all the things that I do, things
2.90 1.145
related to my work would be at or near the top.
I am satisfied with the progress I have made towards meeting my
3.48 .911
goals for the development of new skills.
My chosen line of work gives me a sense of well-being. 3.59 1.211
The most important things that happen to me involve my job. 2.17 .658
My job is a very small part of myself (R) 3.24 .988
I am very personally involved in my job. 3.24 1.123
I live, eat and breathe my job. 1.83 .759
Most of my interests and friends are centred around my job. 2.34 1.010
Most of my life goals are job-centred. 2.28 .996
I like to be absorbed in my job. 3.00 1.134
I consider my job as central to my existence. 2.07 .998
I have very strong ties to my job. 2.90 1.291

Burnout
Burnout was measured using the Maslach Burnout Inventory General Survey. Mean values and standard
deviations for each item are presented in Table 62.

Table 62 Burnout
Mean Std.
Value Deviation
I feel emotionally drained from my work. 2.90 1.520
I have become less interested in my work since I started this job. 2.38 1.801
I have accomplished many worthwhile things in this job. 3.55 1.882
I feel tired when I get up in the morning and have to face another day on the
2.52 1.503
job.
Working all day is really a strain for me. 2.10 1.520
I can effectively solve the problems that arise in my work. 4.28 1.811
I feel burned out from my work. 1.82 1.307
I feel I am making an effective contribution to what my organisation does. 3.96 1.895
I have become less enthusiastic about my job. 2.79 1.897
In my opinion, I am good at my job. 2.04 1.575
I feel exhilarated when I accomplish many worthwhile things in this job. 4.97 1.349
I feel used up at the end of the work day. 4.00 1.764
I just want to do my job and not be bothered. 2.79 1.878
I have become more cynical about whether my work contributes anything. 3.07 2.298
I doubt the significance of my work. 2.79 2.077
At my work, I feel confident that I am effective at getting things done. 2.21 2.024

Respondents were asked to read statements of job-related feelings and decide if they ever felt this way
about their job. Frequency is indicated on a 7-point Likert scale ranging from 0 (never) to 6 (every day).
The 16-item inventory comprises three sub-scales assessing emotional exhaustion, cynicism and

113
professional efficacy. The items for the third dimension of burnout are framed in positive terms and
thus a low score reflects a low sense of professional efficacy.

As burnout is a multi-dimensional, items relating to each construct were summed and a mean value
calculated for emotional exhaustion, professional efficacy and cynicism. These are presented in Table
63 and will be considered separately in further analyses. Internal reliability coefficients for emotional
exhaustion, professional efficacy and cynicism were .85, .87 and .88 respectively, indicating a high level
of internal reliability.

Table 63 Mean score for burnout

Mean Std.
Value Deviation
Emotional exhaustion 2.43 1.201
Professional efficacy 4.16 1.323
Cynicism 2.51 1.639

Work-life control
The degree of control over balancing work with other interests or responsibilities was ascertained via a
three-item 5-point Likert scale (where 1 indicated very little control and 5 indicated very great
control). Of the 29 respondents who worked, six indicated the first item was not applicable to them as
they had no dependants. The average score for the remaining 23 respondents was 3.30. The mean
scores for items 2 and 3 are provided, along with the results for item 1, in the table below. The results
indicate that the partners consider they have a moderate level of control over their work-life balance.
The average score for this scale was calculated by considering the 23 respondents only. The mean score
for the scale was 3.40 (SD=1.059) and the Cronbach alpha for the scale was .97.

Table 64 Work-life control

Mean Std.
Value Deviation
In general, how much control do you have over the way you balance working
3.30 1.222
and dependent care responsibilities?
In general, how much control do you have over the way you balance working
3.34 1.045
and other interests in your life?
In general, how much control do you have over the way you balance working
3.41 1.018
and other responsibilities in your life?

Relationship with spouse/partner

Aspects of the relationship between respondents and their spouse/partner was measured using four
scales: relationship sociability; companionship with partner; relationship tension; and social
undermining. These were also used in the employee survey.

Relationship sociability and companionship with partner

For the first two scales participants were asked to indicate how frequently they and their partner had
undertaken various things together in the past few weeks. A 5-point Likert scale was used ranging from
not at all (0) to very frequently (4).

The items were summed and a mean score obtained.

114
Table 65 Relationship sociability
Mean Std.
Value Deviation
Visited friends together. 1.49 1.140
Entertained friends in your home. 1.21 1.102
Ate out together in a restaurant. 1.65 1.303
Gone out together to a cinema, to play sport or other entertainment. 1.45 1.299

The mean item score for the relationship sociability scale was 1.44 (SD=.967). The Cronbach alpha
coefficient for the scale was .874 indicating an acceptable level of internal consistency reliability. This
indicates quite a low level of relationship sociability.

Table 66 Companionship with partner

Mean Std.
Value Deviation
Had a good laugh together or shared a joke. 2.53 1.039
Been affectionate towards one another. 2.83 1.049
Spent an evening chatting with one another. 2.43 1.037
Did something the other particularly appreciated. 2.66 .984
Taken a drive or walk for pleasure. 1.85 1.351

The mean score for the companionship with partner scale was 2.46 (SD=.856). The Cronbach alpha
coefficient for the scale was .796 indicating an acceptable level of internal consistency reliability. This
indicates a moderate level of companionship with partner.

Relationship tension
The third scale asked participants to indicate how frequently they and their partner had disagreed or
had differences of opinions about specific issues in the past few weeks. Again a 5-point Likert scale was
used ranging from not at all (0) to very frequently (4). A summed score for each of the scales was
then calculated.
Table 67 Relationship tension
Mean Std.
Value Deviation
Being tired 1.36 1.241
Irritating personal habits 1.66 .915
Household expenses 1.21 1.178
Being away from home 1.45 1.248
How to spend leisure time 1.51 1.140
Time spent with friends 1.17 1.129
Your partners job 1.47 1.266
In-laws .87 .947
Not showing love 1.47 1.300

The mean score for the relationship tension scale was 1.35 (SD=.802). The Cronbach alpha coefficient
for the scale was .864 indicating an acceptable level of internal consistency reliability. This indicates
quite a low level of relationship tension.

115
Social undermining
Social undermining was measured by asking questions about the way their partner treated them.
A 3-point Likert frequency scale ranging from never (1) to always (3) was used which indicated the
perceived level of social undermining received. The mean values for each of the items is presented in
Table 68. The mean value for the scale was 1.52 (SD=.324). As the original scale had a low score
indicative of greater social undermining the scale was revered. A high score now indicates a higher level
of perceived social undermining. The Cronbach alpha coefficient for the scale was .86 indicating an
acceptable level of internal consistency reliability.

Table 68 Social undermining

Mean Std.
Value Deviation
Supports and reassures you when you need it. 2.43 .683
Treats you with anger and aggression. 1.49 .505
Shows that he/she cares about you as a person. 2.68 .556
Complicates your life. 1.74 .642
Provides you with useful information or advice when you need it. 2.51 .585
Shows that he/she does not like you. 1.19 .398
Listens to you when you need to talk about things that are very
2.51 .655
important to you.
Makes you feel unwanted. 1.26 .441
Tells you things that strengthen your self confidence. 2.23 .633
Gets on your nerves. 1.77 .476
Understands your way of thinking. 2.23 .476
Criticises you. 1.79 .587
Gives you direct help, i.e. does something for you or gives you what
2.51 .547
is necessary.
Insults you even when he/she does not want to. 1.36 .529
Makes you feel that you can count on him/her. 2.68 .556

Relationship affected by partners work commitments

Participants were asked to indicate to what extent is their relationship was affected by their partners
work commitments. A 5-point Likert scale was used ranging from very little (1) to very great (5).
The mean value for the scale was 2.91 (SD=1.299), indicating that respondents felt that their partners
work was having some effect on their relationship.

Family dependants
Participants were asked to indicate if they had any dependent children, where dependent children were
classified as being under 18 years. Of the 46 respondents, 26 (56.5%) indicated they had dependent
children.

Parenting behaviours
Two series of questions were presented regarding respondent attitudes towards their children, as well as
their views on their partners attitudes towards their children. Both scales utilised a 5-point Likert scale
ranging from never (1) to always (5). Respondents indicated from a list of 16 parenting behaviours
how frequently or infrequently they behaved (or they felt their partners behaved) in that way towards
their child(ren). The mean and standard deviation for each item is presented in Tables 69 and 70.

116
Table 69 Parental attitudes of respondents
Mean Std.
Value Deviation
I encourage my child to talk about the childs troubles. 3.88 1.166
I show sympathy when my child is hurt or frustrated. 4.16 .898
I explode in anger towards my child. 2.04 .611
I am aware of problems or concerns about my child in school. 3.96 1.098
I am responsive to my childs feelings or needs. 4.04 .841
I yell or shout when my child misbehaves. 2.16 .800
I give comfort and understanding when my child is upset. 4.36 .757
I express affection by hugging, kissing and holding my child. 4.12 .881
I argue with my child. 2.12 .833
I have warm and intimate times together with my child. 3.64 1.075
I give praise when my child is good. 4.28 .792
I am easy going and relaxed with my child. 3.80 .707
I tell my child I appreciate what the child tries or accomplishes. 4.20 .764
I show patience with my child. 3.84 .850
I know the names of my childs friends. 4.24 .779
I joke and play with my child. 3.92 .572

Table 70 Respondents views of parental attitudes of their partners

Mean Std.
Value Deviation
He/she encourages our child to talk about the childs troubles. 3.00 1.118
He/she shows sympathy when our child is hurt or frustrated. 3.40 .913
He/she explodes in anger towards our child. 2.48 .823
He/she is aware of problems or concerns about our child in school. 3.20 .957
He/she is responsive to our childs feelings or needs. 3.32 .900
He/she yells or shouts when our child misbehaves. 2.52 .872
He/she gives comfort and understanding when our child is upset. 3.56 1.158
He/she expresses affection by hugging, kissing and holding our
3.56 1.044
child.
He/she argues with our child. 2.48 .963
He/she has warm and intimate times together with our child. 3.28 1.137
He/she gives praise when our child is good. 3.44 .917
He/she is easy going and relaxed with our child. 3.40 .816
He/she tells our child he/she appreciates what the child tries or
3.44 .961
accomplishes.
He/she shows patience with our child. 3.24 .879
He/she knows the names of our childs friends. 2.84 .898
He/she jokes and plays with our child. 3.36 1.114

Utilising the factor structure from the partners scale regarding attitude towards children, the mean of
the two constructs were calculated. Nurture and support of child had items 1, 2, 4, 5, 7, 8, 10, 11, 12,
13, 14, 15, and 16, and verbal anger towards child had items 3, 6, and 9. A higher mean scores indicated
higher nurture and support of, or higher verbal anger towards, the child.

For the respondents parental attitudes to their own parenting behaviour the Cronbach alpha coefficient
was .92 for the nurture items and .76 for verbal anger was indicating an acceptable level of internal

117
consistency reliability. For the respondents views on their partners parental attitudes the Cronbachs
alpha coefficient was .94 for the nurture items and .92for verbal anger again indicating an acceptable
level of internal consistency reliability.

Table 71 shows the mean scores for both the respondents attitudes and their views on their partners
attitudes to parenting.

Table 71 Mean score for parenting attitudes

Mean Std.
Value Deviation
Nurture and support of child 4.03 .624
Verbal anger towards child 2.11 .622
Partners nurture and support of child 3.31 .759
Partners verbal anger towards child 2.49 .823

Household management
A number of questions were asked to ascertain how household tasks were managed.

Time commitment to parental activities

Participants were asked initially about their, and their partners time commitment to taking care of or
doing things with their children things like feeding, dressing, washing, going to places, helping with
homework, disciplining, talking, reading, and driving them to places on work and non-work days.

The mean score and standard deviations for those who indicated they were parents of dependent
children is shown below.
Table 72 Mean score for hours/day on household activities
Mean Std.
Value Deviation
Hours per day when you are working 4.55 2.460
Hours per day when you are not working 8.44 4.022
Hours per day partner commits when they are working 1.77 1.275
Hours per day partner commits when they are not working 4.96 3.304

Typical household tasks

Respondents were then asked to indicate from a list of typical household tasks:

who normally performed that task;

how often they performed the task (where 1 was several times a day and 5 was less than
once per week); and

their time commitment to that task in a typical week.

In particular, typical indoor and outdoor activities were targeted. The results are presented in Tables 73,
74 and 75.

The results for frequency and time commitment are presented with the mean values and standard
deviation for the whole sample, as well as the mean value for just the male and female respondents. The
results for who performs the task is presented with the actual number of responses for the whole
sample as well as number of male and female responses.

118
Table 73 Who undertakes household activities
Total
sample Male Female
N
Prepares meals
Self always 10 0 10
Self usually 23 1 21
Self and partner equally 11 3 8
Partner usually 3 1 2
Partner always 0 0 0
Other 1 0 1

Puts out garbage

Self always 3 2 1
Self usually 9 0 9
Self and partner equally 18 2 15
Partner usually 12 1 11
Partner always 2 0 2
Other 4 0 4

Cleans house
Self always 14 0 14
Self usually 20 1 18
Self and partner equally 11 3 8
Partner usually 0 0 0
Partner always 1 0 1
Other 2 1 1

Shops for groceries

Self always 13 0 13
Self usually 17 2 15
Self and partner equally 12 1 10
Partner usually 3 1 10
Partner always 0 0 0
Other 2 1 1

Mows the lawn

Self always 1 1 0
Self usually 6 1 4
Self and partner equally 4 1 3
Partner usually 10 1 9
Partner always 19 0 19
Other 2 0 2

Takes care of home maintenance/improvements

Self always 0 0 0
Self usually 6 3 3
Self and partner equally 7 1 6
Partner usually 22 1 21
Partner always 9 0 9
Other 1 0 1

It should also be noted that a lower score in Table 74 indicates greater frequency.
Table 74 Mean score for frequency of household activities
Mean Std. Male Female
Value Deviation only Only
Prepares meals 1.64 .870 2.20 1.56
Puts out garbage 3.15 .932 3.20 3.12
Cleans house 2.96 1.179 4.00 2.80
Shops for groceries 3.55 .619 3.00 3.61
Mows the lawn 4.83 .377 4.75 4.84
Takes care of home maintenance/improvements 4.67 .634 3.80 4.78

119
Table 75 Mean score for hours performed in specified task in an average week
Mean Male Female
Std.
Value Average Average
Deviation
(hrs) (hrs) (hrs)
Prepares meals 10.38 5.448 5.20 11.07
Puts out garbage 1.13 .671 1.10 1.13
Cleans house 6.14 4.100 3.00 6.57
Shops for groceries 2.40 .876 3.00 2.34
Mows the lawn 1.37 1.205 1.38 1.35
Takes care of home maintenance/improvements 1.88 2.524 3.40 1.74
TOTAL SPECIFIED HOUSEHOLD DUTIES 23.12 8.758 16.80 24.04

These findings are consistent with previous research, which has found that while men and women both
undertake household tasks their time commitment to these tasks and the tasks they undertake vary
quite considerably. Typically it is women who often bear a larger domestic load and typically undertake
tasks that require daily attention (such as cooking and childcare). Men, whose paid work hours often
exceed those of their domestic partner, spend less time on household chores but often undertake task,
for which the timing is discretionary (i.e. mowing the lawn). It is the relentlessness of the tasks
undertaken by the women, as well as the increased time commitment, that provides an additional
burden on many women.

Figure 71 indicates the hours spent on household and work related tasks for parents and non-parents. It
is clear that parental status affects both the hours worked outside the home as well as hours worked
within the home.

Figure 71 Hours spent on work and household tasks by parental status

40
Hours worked per week
Total hours in typicial
33.27 household tasks per
week
30

25.5
23.57
Mean

20 18.87

10

0
no dependent children one or more dependent children
Parental Status

Satisfaction with sharing of household chores

Respondents were then asked to indicate how satisfied they were with the sharing of the household
chores. A 5-point Likert scale, ranging from 1 (completely dissatisfied) to 5 (completely satisfied) was
used. A higher score was indicative of a greater level of satisfaction. As can be seen in Table 76, the
male respondents were more satisfied with the chore sharing. However, it should be noted that the
sample of males was very small.

120
Table 76 Satisfaction with sharing of household chores
Whole
Std. Male Female
sample
Deviation Mean Mean
Mean
Satisfaction with sharing of chores 3.72 1.280 4.20 3.66

The figure below indicates a gender breakdown of satisfaction with sharing of chores as well as other
factors discussed earlier. It is clear that some gender differences exist. However, these should be viewed
with caution considering the low sample sizes.

Figure 72 Various work-life factors and parental status

5
Satisfaction with sharing of chores
Relationship affected by partners work
4.089 commitments
3.933
4 Work-life balance control

3.214 3.286 3.238

3
2.733
Mean

0
no yes
Dependent children

Schedule compatibility
Respondents where then asked to indicate how compatible their and their partners schedule were.
A 5-point frequency scale was used ranging from 1 (always) to 5 (never) with a lower score indicative of
a greater compatibility. These values indicate a reasonably high level of incompatibility (see Table 77).

Table 77 Schedule compatibility

Mean Std.
Value Deviation
It is difficult for me and my partner to take holidays
3.87 1.154
at the same time
My partners time off work does not match other
3.81 1.116
family members schedules well

121
Bi-variate correlations for spouse/partner

Due the low sample size of 46 it was decided that some correlations would be performed.

Whole sample
Table 78 shows the bi-variate correlations between the work and family related variables measured in
the study. These results are presented under a series of sub-headings (see below). Readers should note
that significant linkages are only described once.

Relationship with partner

There was a significant positive correlation between companionship with partner and relationship
sociability. Not surprisingly, relationship tension was significantly and positively correlated with social
undermining. There were significant negative correlations between the degree to which spouses
perceived their relationship is affected by partner work commitments and relationship factors
(companionship; sociability; and sharing of household chores. In addition, there were significant
positive correlations between social undermining and total household work hours.

Work hours within the house

Spouse/partners of the public and private sector employees participating in the study indicated the
number of hours of work per week undertaken typical household tasks. Number of household work
hours significantly negatively correlated with relationship sociability and partner companionship. Those
who undertook more household work had poorer relationships with their spouses/partners.
Satisfaction in sharing chores was significantly and positively associated with better spousal
relationships.

Table 78 Bi-variate correlations between family and partners work variables for whole
sample
Mean 1 2 3 4 5 6
1. Companionship with partner 2.46 1 .672(**) -.413(**) -.578(**) -.376(**) .506(**)
2. Relationship sociability 1.44 .672(**) 1 -.133 -.334(*) -.428(**) .354(*)
3. Relationship tension 1.35 -.413(**) -.133 1 .566(**) .274 -.412(**)
4. Social undermining 2.48 -.578(**) -.334(*) .566(**) 1 .145 -.448(**)
5. Total household work (hrs) 23.12 -.376(**) -.428(**) .274 .145 1 -.198
6. Satisfaction with sharing of
3.72 .506(**) .354(*) -.412(**) -.448(**) -.198 1
chores
7. Relationship affected by
2.91 -.566(**) -.255 .544(**) .424(**) .449(**) -.467(**)
partner work commitments

** Correlation is significant at the 0.01 level (2-tailed).

* Correlation is significant at the 0.05 level (2-tailed).

Partners who work

The spouses/partners of the public and private sector employees were asked to indicate whether they
were employed. Of those who did, there was a significant negative correlation between cynicism and
hours of work per week. Those who indicated they worked more hours had a lower level of cynicism,
fewer dependent children, and less total household work hours. There was a significant positive
correlation between total household work hours and number of dependent children, and degree to
which their spousal relationship was affected by their partners work commitments. See Table 79.

122
Table 79 Bi-variate correlations between family and partners work variables for
parents within the sample
Mean 1 2 3 4 5 6 7 8
1. Hours worked per week 28.59 1
2. Work involvement 2.71 .350 1
3. Emotional exhaustion 2.43 -.017 -.170 1
4. Cynicism 2.51 -.384(*) -.301 .379(*) 1
-
5. Professional efficacy 4.16 .129 .298 .084 1 .
.532(**)
6. Work-life balance control 3.40 .009 .011 -.174 -.068 ..337 1
7. Relationship affected by
3.00 -.048 .037 .199 .077 .053 -.141 1
partners work commitments
8. Dependent children
0.48 -.394(*) .128 .165 -.017 -.013 -.152 .235 1
(no=0, yes=1)
9. Total household work (hrs) 22.07 -.515(**) -.157 .209 .073 .072 -.093 .406(*) .438(*)

** Correlation is significant at the 0.01 level (2-tailed).

* Correlation is significant at the 0.05 level (2-tailed).

123
124
Section 4: Discussion and
conclusions
Public/private sector differences

Private sector
The results highlight some important differences between the work-life experiences of employees in the
public and private sector construction organisations in our study. These have implications for
management in these two organisations. Employees in the private sector organisation worked longer
hours and experienced more rigid, inflexible work schedules than employees in the public sector.
Compared with public sector employees, private sector employees found it harder to engage in family
and leisure activities. Private sector employees also reported higher levels of work-to-family conflict and
a stronger intention to quit their jobs than employees in the public sector. These results suggest that
construction jobs in the private sector are leaner, necessitating higher workloads (at least in terms of
time involvement) and permitting less flexibility. Batt and Valcour (2003) suggest that two concepts
capture the notion of work-family integration: work-family conflict; and perceived employee control
over managing work and family demands. Our results suggest that private sector employees enjoy
considerably less work-family integration than public sector employees.

The higher levels of work-to-family conflict experienced by private sector employees is of concern,
because work-to-family conflict was found to be a linking mechanism between job demands and family
functioning: in particular, relationship quality with spouse/partner. Thus, in the study, respondents with
higher levels of work-to-family conflict were likely to experience greater relationship tension and lower
relationship quality than employees with lower levels of work-to-family conflict. Work-to-family
conflict also predicted extrinsic job satisfaction and turnover intention in our sample, such that
employees whose levels of work-to-family conflict are high were less satisfied with, and more likely to
quit, their jobs than employees who experienced lower levels of work-to-family conflict. It is possible
that employees in the private sector would eventually withdraw from this sector and move into public
sector jobs as a means to achieve better work-family integration. While the cross-sectional nature of the
study prevents us from determining whether this is actually the case, longitudinal research may be
undertaken in the future to map the career paths of construction professionals and identify the extent
to which family issues impact upon the nature and timing of career choices or decisions: in particular,
their decision to pursue careers in the private or public sectors. Employee perceptions of work-life
balance in private compared to public sector jobs will also be explored in the qualitative interview-based
phase of the current study.

The higher level of turnover intention should be a concern to private sector organizations, given that
the Australian construction industry faces a tight labour market and impending skills shortages. In
particular, the costs of turnover are high for companies employing technical, professional and
managerial employees who are difficult to replace. With dual-earner families now comprising a
significant proportion of the Australian labour market, the implementation of family-friendly work
practices is critical to the attraction and retention of a skilled workforce

Public sector
The high level of continuance commitment among public sector employees is also a concern, because
continuance commitment is consistently linked with lower levels of employee performance and
organisational citizenship behaviour, i.e., employees performing over and above the minimum
requirements of their job. Continuance commitment differs from affective commitment in that with the
former, employees only remain with the organisation because they perceive the costs (either social or

125
economic) of leaving to be too great (Jaros, 1997). One public sector employee expressed this as
follows:

I have no intention of leaving my current employer before I retire. The job is not exciting but I have too much invested in
skills, superannuation, leave, friends etc to consider leaving.

Yousef (2002) expands this definition of continuance commitment to include remaining in an

organisation due to low perceived alternatives (Dunham, Grube, & Castaneda, 1994). Thus employees
remain in their jobs, not because they want to, but because they have to.

This is also consistent with the lower levels of intention to turnover among public sector employees.
The results of our study do not suggest reasons for the high level of continuance commitment among
public sector employees, and further studies are needed to explore its antecedents and consequences. It
is likely that the public sector employee perceptions of being trapped in their jobs is detrimental to
organisational performance.

Insight into public sector employee commitment might be gained from previous research examining
organisational culture in public sector organisations. Parker and Bradley (2000) suggest that, in the
1990s, the public sector in Queensland attempted to depart from the traditional bureaucratic model of
public administration and embark upon a strategy of new public management. A key feature of the
new public management was an attempt to develop more responsive models of management by
introducing performance management and performance-based rewards for managers. They suggest that
public sector organisations also incorporated more participatory and flexible forms of work
organisation in an attempt to engender organisational commitment in employees. Despite this attempt,
Parker and Bradley (2000) reveal that the public sector in Queensland remains bureaucratic, hierarchical
and procedure-driven. Further, they suggest that the ability of public sector managers to pursue
organisational goals is constrained by political considerations and blurred objectives. The political
forces acting upon employees were cited to be a cause of dissatisfaction in the additional comments
provided by a considerable number of public sector employees. For example, one employee wrote:

[I am] sick of the bureaucracy in the organisation. [I am] sick of the lack of fairness [there is] too much political
interference in the process.

The shift to performance-based rewards coupled with political constraints in the public sector has the
potential to create tension for employees. This might account for why public sector employees report a
similar high level of burnout to employees in the private sector organisation. Parker and Bradley (2000)
also suggest that, in the context of the bureaucratic public sector culture, promotion is also likely to be
based on administrative experience rather than entrepreneurial ability. Thus, employees who have
attained managerial status in this sector might perceive that they cannot transfer this experience to work
in the private sector, increasing the perception of being trapped.

An attempt should be made to overcome the high level of continuance commitment in public sector
organisations. This would require a finer-grained analysis of its causes. We suggest that these might be
related to a tension between the new public management ethos, described by Parker and Bradley
(2000), and a traditional public sector focus on politics and stability. This necessitates that any
organisational intervention designed to foster employee cohesion, morale and loyalty needs to take into
account problems already encountered in the management of change in that sector and address the
underlying organisational culture (Beer 1990).

Burnout

Both public and private sector employees reported high levels of burnout. In comparison with
international norm scores for the burnout dimensions, the levels of burnout experienced by
respondents in our sample were high. For example employees in our sample experienced higher levels
of emotional exhaustion than military personnel, technologists and general managers in other studies.
The only occupational group reporting higher levels of emotional exhaustion was nursing. Levels of

126
cynicisms were higher in our sample than those in all other occupational groups, including nurses.
Professional efficacy was also lower in our whole sample than in all other occupational groups. It
should also be noted that on this dimension of burnout public sector employees were significantly
lower than private sector employees.

These findings should be a cause for serious concern for both public and private sector organisations.
Research evidence suggests that burnout is associated with negative outcomes for both individuals and
organisations. For example, at an individual level, burnout has been associated with the experience of
psychological distress, anxiety, depression, reduced self-esteem and unhealthy behaviours, such as
substance abuse (Maslach et al. 2001). Burnout is also understood to be a result of exposure to work-
related stressors and, as such, should be regarded as an occupational health issue.

The link between work, stress and death is receiving more attention and some legal recognition. For
example, a recent study by the Uniting Churchs Urban Ministry Network in Australia revealed that
work pressures were a significant factor in 109 suicides investigated by the Victorian coroner between
1989 and 2000. The report suggests this is likely to be an underestimation because of the lack of detail
required by the coroner about work-related factors. The occupations most commonly affected included
trades (19 per cent), those in supervisory positions (18 per cent) and professionals (14 per cent).
Significantly, in the United Kingdom a widow is reported to have received a compensation settlement
after a court ruled her husbands suicide was the result of work stress (The Age, 16 November, 2002). A
South Australian court also recently linked work stress to bowel cancer, awarding compensation to the
widow of a man who died of the disease (Sydney Morning Herald, 29 March, 2003). These rulings
emphasise the need for employers to address work stress as an occupational health issue and
proactively identify, assess and manage the risks of psycho-social hazards in the work environment.

Failure to prevent or treat employee burnout is also likely to have negative impact on organisational
effectiveness (Wright & Bonett 1997). For example, research indicates that burnout is associated with
absenteeism, turnover, reduced productivity or effectiveness and lower levels of satisfaction and
organisational commitment (Maslach et al. 2001; Schaufeli & Enzmann 1998). Some research suggests
that burnout is contagious, spreading to affect colleagues of those who experience it and even
resulting in a negative spillover into home life (Burke & Greenglass 2001, Cordes & Dougherty 1993,
Westman et al. 2001).

Research suggests that psycho-social hazards arising in the work environment that are associated with
burnout include the following:

Perceived lack of fairness;

Excessive job demands;

Long work hours;

Role conflict,;

Role ambiguity; and

Work-family conflict.

Further research into some of the predictors of burnout is warranted, particularly in the public sector
organisation in which work hours, schedule demands and work-to-family conflict were significantly
lower than in the private sector organisation. It is possible that in the public sector organisation other
sources of burnout exist, for example role conflict and/or role ambiguity arising as a result of the
blurring of organisational goals and political imperatives (see Public sector above).

More recently, the role of the work-family interface as a mediator in the job stressor-burnout
relationship has been evaluated. The Effort-Recovery (E-R) model (Meijman & Mulder, 1998) has been
used to explain the mechanism by which work interference with family life mediates the stressor-strain
relationship (Guerts et al. 2003; Demerouti et al. 2004). According to this model, work-family conflict is
represented as a situation in which employees are unable to recover from the physiological and
psychological by-products of work in the home domain. The inability to recover means that work-
family conflict acts as a linking mechanism between job stress and employee burnout. This effect was
described in greater detail under the heading Work-family conflict in the Introduction to this report.

127
The results of this study are consistent with those of other researchers who report that work-family
conflict mediates the relationship between job stressors and employee burnout (see, for example
Bacharach, et al. 1991 and Guerts et al 2003). This mediation effect was also explored in the pilot study.
Analysis of the Baulderstone Hornibrook data supported the hypothesis that work-to-family conflict
mediates the relationship between job schedule demands and employee burnout (Lingard & Francis, in
press). This replication of results across different samples suggests the hypothesised pattern of
relationships is robust.

The results suggest that the direct relationship often reported between job schedule demands, such as
work hours and irregular hours, and burnout (emotional exhaustion) will be best understood as an
indirect relationship. In this relationship, the perception that work interferes with family or home life
acts as an important intervening variable between job schedule demands and emotional exhaustion.
Thus, job schedule demands may not directly cause emotional exhaustion, but instead give rise to work-
family conflict. In this situation, reduced opportunity to recover during non-work hours leads to
emotional exhaustion.

The implication of these findings is that the reduction of work-to-family conflict, for example, through
the implementation of family-responsive management interventions, may be an effective control
measure for the risk of employee burnout. Emotional exhaustion is the burnout dimension most
strongly and consistently related to negative consequences for both individuals and organisations.
Given the very high levels of burnout in the current sample (among both public and private sector
employees), the urgency with which this issue needs to be addressed is increased.

Supportive work environment

Our results clearly demonstrate the importance of a supportive work environment. Perceived
organisational support (POS), emotional support from ones supervisor, and emotional support from
ones co-workers all demonstrated significant main effects on the emotional exhaustion dimension of
employee burnout. Thus individuals who enjoyed these types of support had lower levels of emotional
exhaustion. Also, where employees perceived their supervisors to be critical or resentful of work-family
issues, emotional exhaustion was also high. In addition to these main effects, our results also suggest
that POS and practical support by supervisors and co-workers for employees mitigates the ill-effects of
work-to-family conflict. Thus, where support is high, the positive relationship between work-to-family
conflict and emotional exhaustion is weaker than where support is lacking. This suggests that these
types of supports can act as a buffer from burnout caused by work-to-family conflict. Further,
employee perceptions that their supervisors were critical or resentful of their family issues also altered
the relationship between work-to-family conflict and emotional exhaustion, such that the relationship
was much stronger where perceived resentment/criticism was high. Our results are consistent with
those of previous researchers (see, for example, Thomas & Ganster, 1995). The results also have
important implications for management.

The results suggest that the creation of a supportive work environment is likely to be an important
preventive strategy for employee burnout and work-family conflict. Employee perceptions of support
can occur at several levels: co-worker (or workgroup); supervisory; and organisational. Allen (2001)
suggests that employees can form different perceptions relating to supportiveness at different levels.
Thus, it is possible that an employee perceives an organisation to be generally supportive of work-life
balance, but perceives his/her immediate supervisor to be lacking in support. In these circumstances,
the benefits of formal work-life balance policies or benefits offered at the organisation level are likely to
be of little or no success.

At an organisational level, there is a need to address organisational norms, values and assumptions
about work-life balance. Research suggests that these norms, values and assumptions can dissuade
employees from using work-life benefits, even when they are provided (see for example, Kirby &
Krone, 2002). Organisational norms regarding the appropriate interaction between work and family can
be shaped by company policies, such as discouraging breakfast meetings or asking questions of anyone

128
who works back late on a regular basis. However, it is also important that company values that are
supportive of work-life balance are clearly communicated in an unambiguous way. For example many
employees believe that organisations desire employees to devote themselves to work at the expense of
other life domains. Where stated values are at odds with the expectations of managers, the false rhetoric
will rapidly lead to cynicism and employee dissatisfaction.

Supervisors play a key role in shaping the work experiences of their subordinates, and often act as
gatekeepers through whom employee requests for work-life balance provisions are processed. For
example, previous studies have suggested that supervisors treat requests by male and/or managerial
employees for flexibility in their work arrangements unfavourably (Barham et al. 2001). Such biases
were not explored in our study but could pose a problem in the traditionally male culture of the
Australian construction industry. Our results emphasise the importance of supervisory support to
employees in our sample and the findings underscore the importance of providing training to
supervisors on how to deal with employees work-life balance issues and the appropriate provision of
flexibility or use of work-life balance benefits.

A particularly strong moderation effect was found for the criticism/resentment dimension of
supervisory support. In our sample, when supervisors were perceived to be critical/resentful of
employees family issues, employee burnout was higher and the relationship between work-to-family
conflict and burnout more pronounced. Similar results were reported by Tepper (2000) who found
abusive supervision to be negatively related to less favourable attitudes towards job, life and
organisation, higher turnover, increased psychological distress, and greater conflict between work and
family life. Particular emphasis in supervisory training should focus on eradicating criticism and
resentment from supervisory behaviours.

The final level upon which support can impact upon employee work-life balance is co-worker support.
In our sample, where co-workers were perceived to be supportive, burnout was lower and the
relationship between work-to-family conflict and burnout was less pronounced. The implication for
organisational interventions is that team-building within work groups is important. Previous research
has sought to identify factors which predict co-workers resentment of employees who use work-life
benefits. Certain groups of workers, for example, older workers and those engaged in jobs requiring a
greater degree of task interdependence, are reported to view work-life balance benefits more favourably
(Parker & Allen, 2001). Hegtvedt et al. (2002) also reported that self-interest (i.e. an individuals own
intended use of a benefit) was not a predictor of resentment, but that gender and gender-role attitudes
were. They report that male employees and employees with traditional gender-role attitudes were more
likely to be resentful of colleagues who use family-friendly policies. This study did not investigate the
extent or correlates of co-worker resentment, however, previous studies suggest that employee training
focusing on the strategic importance to organizations of work-life balance may be needed. Additionally,
the role of gender equity and the need for to accommodate the work-life issues of both men and women
would also be of benefit.

Finally, the different role of emotional versus practical support should be noted. Emotional support
(i.e. listening to employees family issues) from both supervisors and co-workers was associated with
lower levels of emotional exhaustion. However, emotional support did not moderate the relationship
between work-to-family conflict and burnout. Only practical support offered by supervisors or co-
workers weakened the relationship between work-to-family conflict and burnout. The implication of
this finding is that, in order to buffer the effect of work-to-family conflict, supervisors and co-workers
must actively provide practical support to assist employees. This could involve re-arranging or
swapping work shifts and other practical measures.
Flexibility and control

Karasek (as cited in Wall, Jackson, Mullarkey & Parker, 1996), suggested that the level of job strain
experienced by a person is not contingent on level of demands alone, but, is a function of the interaction
between high job demands and low degree of control. In this regard, control is the belief that one can
exert some influence over the environment, either directly or indirectly, so that the environment becomes
more rewarding or less threatening (Ganster & Fusilier, as cited in Thomas & Ganster, 1995, p.7). We

129
examined the role of flexibility or control that employees have over their work and tested the extent to
which this perceived control has a main effect on employees levels of work-to-family conflict and/or
moderates the relationship between work hours and work-to-family conflict. This study found that
perceived control had a main effect on work-to-family conflict, but did not moderate the relationship
between work hours and work-to-family conflict. This finding is similar to that of Behson (2002), who
considered the role of informal work accommodations to family in helping employees to cope with
family-to-work conflict. He suggested that the boundaries between work and family were permeable in
that employees, particularly parents, often attended to family issues while at work, and to work demands
while at home. Behson suggested that employees used a range of informal supports, rather than formal
work-family policies to achieve this. He defined informal work accommodations to family (IWAF) as a
set of behaviours in which employees temporarily adjust their usual work patterns in an attempt to
balance work and family responsibilities. These may alter how, when or where work gets done, but the
work output remains approximately the same. Informal accommodations differ from formal work-family
balance policies, because formal policies create permanent or semi-permanent separations between work
and family, e.g. through job sharing, parental leave, part-time work etc. In another study, Behson (in
press) reports that informal supports of work-life balance have a stronger effect on work-family conflict,
stress, turnover intention and job satisfaction than formal family-friendly benefits. Informal
accommodations are equivalent to our notions of flexibility/control because they relate to job autonomy.
They include such things as leaving work early but taking work home in order to attend to family matters.

Given the significant main effect of flexibility/control on employee work-to-family conflict in the current
study, we suggest that work-to-family conflict could be considerably reduced where organisations foster
work environments conducive to work-family balance. Indeed Behson (2002) suggests that everyday
work flexibility is much more important than more publicized, formal forms of family-friendliness.
However, Behson also notes that managerial support has a significant effect on employees use of
informal accommodations and suggests that the extent to which a supervisor is supportive of employees
work-family balance will influence the employees use of informal means of accommodating family (see
also below). We recommend that employees adaptive behaviours and the role of supervisors support be
investigated in future research. Unless they are abused, informal work accommodations, through
flexibility and control, should not significantly affect the quantity or quality of work an employee
accomplishes but could considerably reduce the negative outcomes of work-to-family conflict, benefiting
individuals, their families and organisations.

Work-to-family and family-to-work conflict

In this study, work-to family conflict was positively and significantly related to employee turnover
intentions. This finding is consistent with the results of previous research (Boyar et al. 2003). The
finding has important implications for human resource managers who might wish to reduce the
economic and personal costs associated with employee turnover. Efforts to reduce work-to-family
conflict through the creation of policies and procedures supportive of work-life balance, or minimising
work-family conflict by addressing specific antecedents, such as work hours or subjective quantitative
overload, is likely to be beneficial to organisations. Our results suggest that job schedule demands,
especially work hours, as well as employees subjective sense of workload have the potential to
indirectly influence turnover intentions through their effect on work-to-family conflict. In an effort to
reduce subjective quantitative overload, time management training could be provided to employees to
help them to utilise their work hours effectively. Additionally, companies need to be realistic about the
expectations placed on employees.

The integrative model of the work-family interface developed indicated that the non-work domain is
more susceptible to interference from work than the work domain is to interference from family. Thus
work-to-family conflict had a negative impact on family relationships, in particular, the relationship with
ones spouse/partner. However, work attitudes and/or distress were unaffected by family-to-work
conflict. This asymmetrical pattern of relationships suggests that employees in our sample were better
at segregating and protecting the work domain from family interference than vice-versa. It also
suggests that these employees were more likely to make adjustments in their non-work lives to
accommodate work demands than vice versa. This may be due to the fact that adjustments to family life

130
are less likely to impact upon their livelihood. Assumptions about the relative importance of work and
family to employees in construction organisations should be explored in future studies to determine
whether role salience or the value placed by individuals on family vis--vis work plays a role in shaping
relationships at the work-family interface.

Work-family preferences

Preferences for work-life benefits were found to vary considerably between different groups of
employees in our sample. Childcare assistance support was highest among employees aged 30-49 years
and was significantly greater among employees who were partnered with dependent children than those
in other family structures. By comparison, wellness and personal development benefits were highest
among the 20-29 year age group and progressively declined with age. They also most strongly preferred
by single persons and partnered employees without dependent children. The results suggest that no one
work-life balance solution is applicable to all employees and therefore workforce profiling by business
units or organisational sub-units might be a useful means of identifying which initiatives to implement.
It is likely that a cafeteria style range of benefits might need to be offered from which employees can
design a salary and benefit package that best suits their needs at a particular life-stage. The key is likely
to be in affording employees a sense of flexibility and control over their work arrangements. It is also
important that organisations consider their public face and appeal. The expectations of these workers
may not be the same as those of existing employees, especially in an industry in which considerable
segregation by gender still exists. We therefore recommend that both organisations consider the nature
of their workforce carefully prior to deciding upon work-life balance strategies, in particular formal
benefits to be offered. .

Recommendations

Public and Private sector organisations

Both organisations need to examine strategies and initiatives to reduce employee burnout. At present it
is likely that employees are under-performing and possibly suffering adverse behavioural or health
effects due to the extreme levels of burnout. Known antecedents of burnout include:

Work hours;

Subjective workload;

Role conflict; and

Work-family conflict.

Interventions designed to reduce work-family conflict could be particularly beneficial, because job
demands affect burnout via work-to-family conflict. General interventions could include:

Strategies to improve employees sense of control over their workplace arrangements (variable
had a main effect on work-to-family conflict).

Strategies to improve the extent to which employees can use informal supports, for example,
taking work home in order to meet family obligations;

Strategies to reduced work hours, for example alternate Saturdays off for project-based staff;
and/or

Strategies to publicise the availability and uptake of family-friendly policies by employees of
either gender, such as profiling male employees who have utilised work-life benefits in
newsletters etc.

However, the results of this study also indicate that employee preferences for work-life balance
initiatives will depend upon their age and life-stage. Thus, quick-fix solutions for all employees cannot
be specified. Within organisational business units or sub-units, workforce profiling should be
undertaken to ensure that strategies meet the needs of constituent groups of employees.

131
Care should also be taken to consider strategies that may not be a priority for existing employees, but
which might be implemented in order to improve workforce diversity and attract groups of employees
who are not currently well represented.

Both organisations need also to focus on creating a work environment that is supportive of work-life
balance. In particular, strategies to be developed:

Create a supportive organisational culture;

Coach supervisors in supporting employees work-life balance; and

Foster workgroups that are supportive of employee work-life balance.

Aspects of the organisational culture to be addressed include:

Ensuring that employees do not perceive negative career consequences associated with using
work-life benefits;

Ensuring that gender equity in the availability and use of benefits is encouraged; and

Evaluating and, if necessary, changing shared assumptions about the appropriate relationship
between work and non-work life.

Supervisors in the traditional and male dominated construction environment are likely to require
training to enable them to respond supportively to subordinates work-life balance issues.

The strategic importance of work-life balance to organisations should be communicated to all

employees in order to foster positive attitudes to work-life balance within work groups.

Public sector
A main concern to the public sector organisations is the high level of continuance commitment, which
is likely to be detrimental to organisational performance. While this study did not explore the
antecedents of continuance commitment, work-to-family conflict negatively related to continuance
commitment. Thus as conflict increases, continuance commitment decreases. Further research is
necessary in this area. It is recommended that in the first instance the public sector organisation
conduct further research to explore why the high level of continuance commitment exist. This can be
could be explored in the qualitative component of the current study.

Respondent comments suggest that public sector employees might be dissatisfied due to frustration
with the bureaucratic nature of the organisation and a sense that they are constrained by procedures
and prevented from doing their job. The gap they perceive between the administrative experience
required to be promoted in the public sector and the entrepreneurial ability required in the private
sector might make these relatively senior employees feel trapped.

It is suggested that the public sector organisation examine possibilities for:

Job enrichment or rotation, to ensure employees perceive that their skills and experience
remain current;

Strategies to support managerial autonomy and decision-making; and

Professional development opportunities to ensure employees technical and managerial skills
remain up-to-date and in line with practices in the private sector.

Private sector
The private sector organisations should address the following issues:

Employees work-to-family conflict as a means of reducing turnover intentions;

Develop strategies to improve employees control over work place arrangements;

Consider ways to reduce employees work hours; and

Develop a work culture more conducive to work-life balance.

The private sector could learn important lessons from the public sector in all of these areas. However,
it must be acknowledged that time and cost pressures arising from the competitive tendering process

132
can present serious difficulties for these strategies. An organisation-wide, rather than a single project-
focus, must be taken to the issues of work-life balance. There is emerging empirical evidence to support
the bottom line benefits of work-life balance. Moreover, construction industry clients should also be
made aware that projects are more likely to run smoothly and be completed on-time when the
contractors workforce is satisfied and balanced.

133
134
References
Aldous, J., Osmond, M.W., & Hicks, M.W. (1979). Mens work and mens families, in W.R. Burr, R.
Hill, F.I. Nye, & I.L. Reiss (Eds), Contemporary Theories about the Family. New York: The Free Press.
Allen, N. J. & Meyer, J. P., (1990). The measurement and antecedents of affective, continuance and
normative commitment to the organization. Journal of Occupational Psychology, 63, 1 - 18.
Allen, T. D., (2001). Family-supportive work environments: The role of organizational perceptions.
Journal of Vocational Behavior, 58, 414 - 435.
APESMA (2000). APESMA working hours and employment security survey report, PowerPoint presentation,
(personal communication).
Appels, A. & Schouten, E. (1991). Burnout as a risk factor for coronary heart disease. Behavioral
Medicine, 17, 53 59.
Australian Bureau of Statistics. (1994). Focus on families: Work and family Responsibilities, Cat No 4422.0,
Canberra. Australian Government Printing Service.
Australian Bureau of Statistics. (1998). Labour force, Australia, Cat No 6203.0, May, Canberra. Australian
Government Printing Service.
Bacharach, S.B., Bamberger, P., & Conley, S. (1991). Work-home conflict among nurses and engineers:
Mediating the impact of stress on burnout and satisfaction at work, Journal of Organisational Behavior, 12,
39 - 53.
Bardoel, E. A., Moss, S. A., Smyrnios, K. & Tharenou, P. (1999). Employee characteristics associated
with the provision of work-family policies and programs. International Journal of Manpower, 20, 563 - 576.
Barham, L. J., Gottlieb, B. H., & Kelloway, E. K., (2001). Variables affecting managers willingness to
grant alternative work arrangements. The Journal of Social Psychology, 138, 291 - 302.
Barnett, R.C., Garies, K.C., & Brennan, R. T. (1999). Fit as a mediator of the relationship between work
hours and burnout. Journal of Occupational Health Psychology, 4, 307 - 317.
Baron, R.M. & Kenny, D.A., (1986). The moderator-mediator variable distinction in social
psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social
Psychology, 51, 1173 - 1182.
Batt, R. & Valcour, P.M. (2003). Human resources practices and predictors of work-family outcomes
and employee turnover. Industrial Relations, 42, 189 - 220.
Baxter, J. (1993). Work at home: The domestic division of labour. Australia: University of Queensland Press.
Benbow, S.M. & Jolley, D.J. (2002). Burnout and stress amongst old age psychiatrists. International
Journal of Geriatric Psychiatry, 17, 710 - 714.
Beer, M., (1990), Why organizational change programs dont produce change, Harvard Business Review,
68, 158-166.
Biggs, S. (1998). Work practices: The reality versus the rhetoric. Changing families, challenging futures, paper
presented at the 6th Australian Institute of Family Studies conference in Melbourne, 25 - 27 November.
Blair-Loy, M. & Wharton, A.S., (2002). Employees' use of work-family policies and the workplace
social context. Social Forces, 80, 813 - 845.
Bohen, H.H. & Viveros-Long, A. (1981). Balancing jobs and family life: Do flexible work schedules help?
Philadelphia: Temple University Press.
Boise, L. & Neal, M. B. (1996). Family responsibilities and absenteeism: Employees caring for parents
versus employees caring for children. Journal of Managerial Issues, 8, 218 238.

135
Boles, J.S., Howards, W.G., & Donofrio, H.H. (2001). An investigation into the inter-relationships of
work-family conflict, family-work conflict and work satisfaction. Journal of Managerial Issues, 13, 376 -
390.
Bourke, J. (2000). Corporate women, children, careers and workplace culture: The integration of flexible work practices
into the legal and finance professions. Studies in Organisational Analysis and Innovation Monograph Number
15, Industrial Relations Research Centre, University of New South Wales. Sydney.
Boyar, S.L., Maertz, C.P. (jr), Pearson, A.W., & Keough, S. (2003). Work-family conflict: A model of
linkages between work and family domain variables and turnover intentions, Journal of Managerial Issues,
15, 175 - 190.
Bruck, C.S., Allen, T.D., Spector, P.E., (2002). The relation between work-family conflict and job
satisfaction: A finer-grained analysis, Journal of Vocational Behavior, 60, 336 - 353.
Butler, A., Gasser, M., & Smart, L., (2004). A social-cognitive perspective on using family-friendly
benefits, Journal of Vocational Behavior, 65, 57 - 70.
Casper, W.J., Martin, J.A., Buffardi, L.C., & Erdwins, C.J. (2002). Work-family conflict, perceived
organizational support and organizational commitment among employed mothers. Journal of Occupational
Health Psychology, 7, 99 - 108.
Chang, E. (1999). Career commitment as a complex moderator of organizational commitment and
turnover intention. Human Relations, 52, 1257 1278.
Cohen, J., Cohen, P., West, S.G., & Aiken, L.S. (2003). Applied multiple regression/correlation analysis for the
behavioural sciences. Lawrence Erlbaum Associates, Mahwah, New Jersey.
Cook, J.D., Hepworth, S.J., Wall, T.D., & Warr, P.B. (1981). The experience of Work: A compendium
and review of 249 measures and their use.
Corcoran, K. (1995). Measuring burnout: An updated reliability and convergent validity study. In R.
Crandall & P.L. Perrewe (Eds), Occupational stress: A handbook. Washington: Taylor & Francis.
Crouter, A. C., Bumpus, M. F., Maguire, M. C., & McHale, S. M.. (1999). Linking parents work
pressure and adolescents well-being: Insights into dynamics in dual-earner families, Developmental
Psychology, 35, 1453 - 1461.
Crouter, A.C., Bumpus, M.F., Head, M.R., & McHale, S. M. (2001). Implications of overwork and
overload for the quality of mens family relationships, Journal of Marriage and the Family, 63, 404 416.
Demerouti, E., Bakker, A.B., & Bulters, A J., (2004). The loss spiral of work pressure, work-home
interference and exhaustion: Reciprocal relations in a three-wave study, Journal of Vocational Behavior, 64,
131 - 149.
Dolan, S.L. (1995). Individual, organizational and social determinants of managerial burnout:
Theoretical and empirical update. Crandall, Rick (Ed); Perrewe, Pamela L. (Ed). (1995). Occupational
stress: A handbook, (pp. 223 - 238). Philadelphia, PA, US: Taylor & Francis.
Dunham, R.B., Grube, J.A., & Castaneda, M.B. (1994). Organizational commitment: The utility of an
integrative definition. Journal of Applied Psychology, 79, 370 380.
Eaton, S. C., (2003). If you can use them: Flexibility policies, organizational commitment and perceived
performance, Industrial Relations, 42, 145 - 167.
Eisenberger, R., Fasolo, P., & Davis-LaMastro, V. (1990) Perceived organizational support and
employee diligence, commitment and innovation, Journal of Applied Psychology, 75, 51 - 59.
Eisenberger, R., Huntington, R., Hutchison, S., & Sowa, D. (1986). Perceived organizational support.
Journal of Applied Psychology, 71, 500 507.
Eisenberger, R., Jones, J.R., Aselage, J., & Sucharski, I.L., (to be published in 2004). In J. Coyle-
Shapiro, L. Shore, & S. Taylor, & L. Tetrick (Eds.). The employment relationship: Examining psychological and
contextual perspectives. Oxford University Press.

136
Fallon, B. & Mallamace, J. (2000). The need for and the availability of family friendly programs does
not mean they will be used. Personal communication to principal authors.
Francis V E (2004) Supportive organisational cultures and their effect on male civil engineers, The
Australian Journal of Construction Economics and Building. Vol4 (1), pp.1 - 9.
Francis V E (2003) Civil engineers and work-family conflict: The role of workplace support on their
work and non-work satisfaction and well-being. Proceedings of the Second International Conference on
Construction in the 21st Century (CITC-II) Sustainability and Innovation in Management and Technology, Hong
Kong, 1012 December 2003, pp.324330.

Francis V E and Lingard H (2002) The case for family-friendly work practices in the Australian
Construction Industry. The Australian Journal of Construction Economics and Building, Vol.2(1), pp.2836.
Frone, M. R., Russell, M. & Cooper, M. L., (1992) Antecedents and outcomes of work-family conflict:
Testing a model of the work-family interface, Journal of Applied Psychology, 77, 65 - 78.
Frone, M.R., (2000), Work-family conflict and employee psychiatric disorders: The national
comorbidity survey, Journal of Applied Psychology, 85, 888 - 895.
Frone, M.R., Yardley, J.K., & Markel, K.S. (1997). Developing and testing an integrative model of the
work-family interface, Journal of Vocational Behavior, 50, 145 - 167.
Gaines, J. & Jermier, J.M. (1983). Emotional exhaustion in a high stress organization. Academy of
Management Journal, 26, 567 586.
Gmelch, W.H. & Gates, G., (1998). The impact of personal, professional and organizational
characteristics on administrator burnout. Journal of Educational Administration, 36, 146 - 159.
Goff, S.J., Mount, M.K., & Jamison, R.L. (1990). Employer supported child care, work/family conflict
and absenteeism: a field study, Personnel Psychology, 43, 793 - 809.
Grandey, A. (2001). Family-friendly policies: Organizational justice perceptions of need-based
allocations. In R. Cropanzano (Ed.). Justice in the workplace: From theory to practice, vol. 2 (pp. 145 - 174).
Mahwah, NJ: Erlbaum.
Grant-Vallone, E.J. & Donaldson, S.I., (2001). Consequences of work-family conflict on employee
well-being over time. Work & Stress, 15, 214 - 226.
Gray, M. & Tudball, J. (2002). Family-friendly work policies: Differences within and between workplaces.
Research report number 7 Australian Institute of Family Studies.
Greenhaus, J.H., & Beutell, N.J. (1985). Sources and conflict between work and family roles. Academy of
Management Review, 10, 76 88.
Greenhaus, J.H., Collins, K.M., & Shaw, J.D. (2003). The relation between work-family balance and
quality of life. Journal of Vocational Behavior, 63, 510 - 531.
Greenhaus, J.H., Parasuraman, S., Granrose, C.S., Rabinowitz, S., & Beutell, N.J., (1989). Sources of
work-family conflict among two-career couples. Journal of Vocational Behavior, 34, 133 - 153.
Greenhaus, J.H., Parasuraman, S.J. & Collins, K.M. (2001). Career involvement and family involvement
as moderators of relationships between work-family conflict and withdrawal from a profession. Journal
of Occupational Health Psychology, 2, 91 - 100.
Grover, S.L. & Crooker, K.J. (1995). Who appreciates family-responsive human resource policies: The
impact of family-friendly policies on the organizational attachment of parents and non-parents Personnel
Psychology, 48, 271 288.
Grzywacz, J.G. & Marks, N.F. (2000a). Family, work, work-family spillover and problem-drinking
during midlife. Journal of Marriage and the Family, 62, 336 - 348.
Grzywacz, J.G. & Marks, N.F., (2000b). Reconceptualizing the work-family interface: An ecological
perspective on the correlates of positive and negative spillover between work and family, Journal of
Occupational Health Psychology, 5, 111 - 126.

137
Guerts, S.A.E. & Demerouti, E. (2003). Work/non-work interface: A review of theories and findings.
In M. J. Schabracq, J. A. M. Winnubst, and C. L. Cooper (Eds), Handbook of work and health psychology.
Chichester: Wiley. Pp. 279 - 312.
Guerts, S.A.E., Kompier, M.A. J., Roxburgh, S., & Houtman, I.L.D. (2003). Does work-home
interference mediate the relationship between workload and wellbeing? Journal of Vocational Behavior, 63,
532 - 559.
Guerts, S.A.E., Rutte, C., & Peeters, M., (1999). Antecedents and consequences of work-home
interference among medical residents. Social Science and Medicine, 48, 1135 - 1148.
Hammer, L.B., Allen, E., & Grigsby, T.D. (1997). Work-family conflict in dual earner couples: Within
individual and crossover effects of work and family. Journal of Vocational Behavior, 50, 185 - 203.
Hammer, T.H., Saksvik, P.., Nytr, K., Torvatn, H., & Bayazit, M. (2004). Expanding the
psychosocial work environment: Workplace norms and work-family conflict as correlates of stress and
health. Journal of Occupational Health Psychology, 9, 83 - 97.
Hegtvedt, K.A., Clay-Warner, J., & Ferrigno, E.D., (2002). Reactions to injustice: Factors affecting
workers resentment toward family-friendly policies. Social Psychology Quarterly, 65, 386 - 400.
Herscovitch, L. & Meyer, J.P. (2002). Commitment to organizational change: Extension of a three-
component model. Journal of Applied Psychology, 87, 474 487.
Hill, E.J., Hawkins, A.J., Ferris, M., & Weitzman, M. (2001). Finding and extra day a week: The positive
influence of perceived job flexibility on work and family life balance. Family Relations, 50, 49 - 58.
Hobfall, S.E. (2001). The influence of culture, community and the nested-self in the stress process:
Advancing conservation of resources theory. Applied Psychology: An International Review, 50, 337 - 421.
Hobson, C.J., Delunas, L, & Kesic, D. (2001). Compelling evidence of the need for corporate work/life
balance initiatives: Results from a national survey of stressful life-events. Journal of Employment Counseling,
38, 38 44.
Hughes, R.E., (2001). Deciding to leave but staying: teacher burnout, precursors and turnover.
International Journal of Human Resource Management, 12, 288 - 298.
Ingram, P., and Simons, T., (1995). Institutional and resource dependence determinants of
responsiveness to work-family issues. Academy of Management Journal, 38, 1466 1482.
Jaros, S.J. (1997). An assessment of Meyer and Allen's [1991] three-component model of organisational
commitment and turnover intentions. Journal of Vocational Behavior, 51, 319 337.
Kamerman, S.B. and Kahn, A.J. (1987). The responsive workplace. Columbia University Press, New York.
Kanungo, R. (1982). Measurement of job and work involvement. Journal of Applied Psychology, 67, 341 -
349
Kelloway, E.K., Gottlieb, B.H., & Barham, L. (1999). The source, nature and direction of work and
family conflict: A longitudinal investigation. Journal of Occupational Health Psychology, 4, 337 - 346.
Kirby, E.L. & Krone, K.J. (2002). The policy exists but you can't really use it: Communication and
the structuration of work-family policies. Journal of Applied Communication Research, 30, 50 77.
Kossek, E.E. & Ozeki, C., (1998). Work-family conflict, policies and the job-life satisfaction
relationship: A review and directions for organizational behaviour-human resources research. Journal of
Applied Psychology, 83, 139 - 149.
Kossek, E.E., Colquitt, J.A., & Noe, R.A. (2001). Care-giving decisions, well-being and performance:
The effects of place and provider as a function of dependent type and work-family climate. Academy of
Management Journal, 44, 29 - 44.
Lambert, S.J. (2000). Added benefits: The links between work-life benefits and organizational
citizenship behavior. Academy of Management Journal, 43, 801 815.
Lewis, V., Hand, K., & Tudball, J. (2001). Family and work: the familys perspective. Department of Family
and Community Services: Canberra, ACT.

138
Lingard, H. (2004). Work and family sources of burnout in the Australian engineering profession: a
comparison of respondents in dual and single earner couples, parents and non-parents. American Society
of Civil Engineers, Journal of Construction Engineering and Management, 130 (2), 290 - 298.
Lingard H and Francis V E (2002) Work-life issues in the Australian Construction Industry: Findings of a pilot
study. Construction Industry Institute, Australia, Research Report, Brisbane, 94pp.
Lingard, H. & Francis, V. (in press).The work-life experiences of office and site-based employees in the
Australian construction industry. Construction Management and Economics, accepted 26th April 2004.
Lingard, H. & Lin, J. (in press). Career, family and work environment determinants of organisational
commitment among women in the Australian construction industry. Construction Management and
Economics, accepted 19 June 2003.
Lingard, H. & Sublet, A. (in press). The impact of job and organisational demands on marital and
relationship quality among Australian civil engineers, submitted to Construction Management and Economics.
Lingard, H. & Sublet, A., (2002). The impact of job and organisational demands on marital and
relationship quality among Australian civil engineers. Construction Management and Economics, 20, 507
521.
Lingard, H., (2003). The impact of individual and job characteristics on burnout among civil engineers
in Australia and implications for employee turnover. Construction Management and Economics, 21, 69 - 80.
Major, V.S, Klein, K.J., & Ehrhart, M.G. (2002). Work time, work interference with family and
psychological distress. Journal of Applied Psychology, 87, 427 436
Marks, S.R; & MacDermid, S.M. (1996). Multiple roles and the self: A theory of role balance. Journal of
Marriage and the Family, 58, 417 432.
Maslach, C., Jackson, S.E. & Leiter, M.P. (1996). Maslach Burnout Inventory Manual. Palo Alto, CA:
Consulting Psychologists Press, 3rd Edition.
Maslach, C., Schaufeli, W.B., & Leiter, M.P. (2001). Job burnout. Annual Review of Psychology, 52, 397 -
422.
Mauno, S. & Kinnunen, U. (1999). The effects of job stressors on marital satisfaction in Finnish dual-
earner couples. Journal of Organizational Behavior, 20, 879 - 895.
McKeen, C.A. & Burke, R.J. (1994). The woman-friendly organisation: Initiatives valued by managerial
women. Employee Counselling Today, 6, 18 - 25.
Meijman, T.F., Mulder, G. (1998). Psychological aspects of workload. In P.J.D. Drenth, H. Thierry, et
al (Eds.), Handbook of work and organizational, Vol. 2: Work psychology (2nd ed.). (pp. 5-33). Hove,
England: Psychology Press/Erlbaum.
Meyer, J.P., Paunonen, S.V., Gellatly, I.R., Goffin, R.D., & Jackson, D.N., (1989). Organizational
commitment and job performance: Its the nature of the commitment that counts. Journal of Applied
Psychology, 74, 152 - 156.
Meyer, J.P., Allen, N.J., & Smith, C.A. (1993). Commitment to organizations and occupations:
Extension and test of a three component conceptualization. Journal of Applied Psychology, 78, 538 - 551.
Meyer, J.P., Stanley, D.J., Herscovitch, L., & Topolnytsky, L. (2002). Affective, continuance, and
normative commitment to the organization: A meta-analysis of antecedents, correlates and
consequences. Journal of Vocational Behavior, 61, 20 52.
Moen, P. & Yu, Y., (2000). Effective work/life strategies: Working couples, work conditions, gender
and life quality. Social Problems, 47, 291 - 326.
Moen, P., Harris-Abbott, D., Lee, S., & Roehling, P. (1999). The Cornell couples and careers study.
Report, Sloan centre for the study of working families, New York: Cornell University.
Netemeyer, R.G., Boles, J.S., & McMurrian, R. (1996). Development and validation of work-family
conflict and family-work conflict scales. Journal of Applied Psychology, 81, 400 - 410.

139
Norusis, M.J. (1993). Professional Statistics. SPSS for Windows, Release 6, SPSS Inc., Chicago.
ODriscoll, M.P., Poelmans, S., Kalliath, T., Allen, T. D., Cooper, C. L., and Sanchez, J. L. (2003),
Family-responsive interventions, perceived organizational and supervisor support, work-family conflict
and psychological strain, International Journal of Stress Management, 10, 326-344.
Orden, S.R. & Bradburn, N.M. (1968). Dimensions of marriage happiness. The American Journal of
Sociology, 73, 715 731.
Orthner, D.K. & Pittman, J.F. (1986). Family contributions to work commitment. Journal of Marriage and
the Family, 48, 573 - 581.
Parasuraman, S. & Simmers, C.A. (2001). Type of employment, work-family conflict and wellbeing,
Journal of Organizational Behavior, 22, 551 - 568.
Parasuraman, S. (1982). Predicting turnover intentions and turnover behavior: A multivariate analysis.
Journal of Vocational Behavior, 21, 111 - 21.
Parasuraman, S., Greenhaus, J.H., & Granrose, C.S. (1992). Role stressors, social support and well-
being among two-career couples. Journal of Organizational Behavior. 13, 339 - 356.
Parasuraman, S., Purohit, Y.S., Godshalk, V.M., & Beutell, N.J. (1996). Work and family variables,
entrepreneurial career success and psychological wellbeing. Journal of Vocational Behavior, 48, 275 - 300.
Parker, R. and Bradley, L., (2000), Organisational culture in the public sector: evidence from six
organisations, The International Journal of Public Sector Management, 13, 125-141
Presser, H.B. (2000). Non-standard work schedules and marital instability, Journal of Marriage and the
Family, 62, 93 - 110.
Robinson, C., Mandleco, B., Olsen, S., and Hart, C. (1995). Authoritative, authoritarian and permissive
parenting practices: Development of a new measure. Psychological Reports, 77, 819 830.
Rothausen, T. (1999). Family in organizational research: A review and comparison of definitions and
measures. Journal of Organizational Behaviour, 20, 817 836.
Russell, G., Barclay, L., Edgecombe, G., Donovan, J., Habib, G., Callaghan, H., & Pawson, Q. (1999).
Fitting fathers into families: men and the fatherhood role in contemporary Australia. Department of Family and
Community Services: Canberra.
Saltzstein, A., Ting, Y., & Saltzstein, G. (2001). Work-family balance and job satisfaction: The impact
of family-friendly policies on attitudes of federal government employees. Public Administration Review,
61, 452 467.
Schaufeli, W. & Enzman, D. (1998). The burnout companion to study and practice: A critical analysis. Taylor &
Francis: London.
Scheibl, F., & Dex, S. (1998). Should we have more family-friendly policies? European Management
Journal 16, 586 599.
Secret, M. (2000). Identifying the family, job and workplace characteristics of employees who use work-
family benefits. Family Relations, 49, 217 - 225.
Selder, F. E. & Paustian, A. (1989). Burnout: Absence of vision. In D.T. Wessells, A.H. Kutscher, I.B.
Seeland, F.E. Selder, D.J. Cherico, & E.J. Clark (Eds.). Professional Burnout in Medicine and the Helping
Professions. USA: Haworth Press, Inc.
Shinn, M. (1982). Methodological issues: Evaluating and using information. In W.S. Paine, (Ed), Job
stress and Burnout. Newbury Park: Sage.
Smithers, G. (2000). The effect of the site environment on motivation and de-motivation of
construction professionals. Proceedings of the 16th Annual Conference of the Association of Researchers in
Construction Management, September 6 - 8, Glasgow Caledonian University, Glasgow, 455-464.
Stanton, J. M., (1998). An empirical assessment of data collection using the internet, Personnel Psychology,
51, 709 - 725.

140
Stevens, J. (1996). Applied Multivariate statistics for the social sciences (3rd ed). New Jersey: Lawrence
Erlbaum Associates.
Stewart, W. & Barling, J. (1996). Fathers work experiences affect childrens behaviours via job-related
affect and parenting behaviours. Journal of Organizational Behavior, 17, 221 - 232.
Sullivan, I.G. (1989). Burnout: A study of a psychiatric centre. In D.T. Wessells, A.H. Kutscher, I.B.
Seeland, F.E. Selder, D.J. Cherico, & E.J. Clark (Eds.). Professional Burnout in Medicine and the Helping
Professions. USA: Haworth Press, Inc.
Sydney Morning Herald, March 29, 2003, Work and stress: Judge finds a deathly link, John Fairfax
Publications Pty Ltd., p. 27.
Tanner, L., (2003), Crowded Lives, Pluto Press, Melbourne.
Tausig, M. & Fenwick, R., (2001). Unbinding time: Alternate work schedules and work-life balance.
Journal of Family and Economic Issues, 22, 101 - 119.
Taylor, J.C. & Bowers, D.G. (1972). Survey of organisations: A machine scored standardized questionnaire
instrument. Ann Arbor, Centre for Research on the Utilization of Scientific Knowledge, University of
Michigan.
Tennant, C. (1996). Experimental stress and cardiac function. Journal of Psychosomatic Research, 40, 569
583.
The Age, June 26, 2003, p. 12, Balancing work and family benefits us all, The Age Company Ltd,
Melbourne.
The Age, May 31, 2003, p. 3, To have a family or not? Its time to get personal, The Age Company Ltd,
Melbourne.
The Age, November 16, 2002, When work is a killer, The Age Company Ltd., p. 40. Melbourne.
The Australian, June 25, 2003, p4, Test case push for working parents, Nationwide News Pty Limited,
Sydney.
Thomas, L.T. & Ganster, D.C. (1995). Impact of family-supportive work variables on work-family
conflict and strain: A control perspective. Journal of Applied Psychology, 80, 6 15.
Thompson, C.A., Beauvais, L.L., & Lyness, K.S. (1999). When work-family benefits are not enough:
The influence of work-family culture on benefit utilization, organizational attachment and work-family
conflict. Journal of Vocational Behavior, 54, 392 - 415.
Thompson, L.F., Surface, E.A., Martin, D.L., & Sanders, M.G. (2003). From paper to pixels: moving
personnel surveys to the web. Personnel Psychology, 56, 197 - 225.
Thornwaite, L. (2002). Work-family balance: International research on employee preferences. Working
paper 79, from the Working time today conference, 16 August, 2002. Sydney: University of Sydney.
Tucci, J., Goddard, C., & Mitchell, J., (2004). The concerns of Australian parents, Australian Childhood
Foundation, Sydney.
Wallace, J. E., (1997), It's about time: a study of hours worked and work spillover among law firm lawyers,
Journal of Vocational Behavior, 50, 227-248.
Watkins, K.E. (1995). Changing managers defensive reasoning about work/family conflicts. Journal of
Management Development, 14, 77 - 88.
Westman, M. & Eden, D. (1997). Effects of respite from work on burnout: vacation relief and fade-out.
Journal of Applied Psychology, 82, 516 - 527.
Westman, M. & Etzion, D. (1995). Crossover of stress, strain and resources from one spouse to
another. Journal of Organizational Behavior, 16, 169 181.
Westman, M., Etzion, D., & Danon, E. (2001). Job insecurity and crossover of burnout in married
couples. Journal of Organisational Behavior, 22, 467 - 481.

141
White, L. & Keith, B. (1990). The effect of shift work on the quality and stability of marital relations.
Journal of Marriage and the Family, 52, 453 - 462.
Williams, K.J. & Alliger, G.M., (1994). Role stressors, mood spillover and perceptions of work-family
conflict in employed parents. Academy of Management Review, 37, 837 - 868.
Wright, T.A. & Bonnett, D.G. (1997). The contribution of burnout to work performance. Journal of
Organizational Behavior, 18, 491 - 499.
Yoon, J. & Thye, S.R. (2002). A dual process model of organisational commitment: Job satisfaction
and organisational support. Work and Occupations, 29, 97 124.

142