2009 Fossetal HRM PDF

ENCOURAGING KNOWLEDGE
SHARING AMONG EMPLOYEES:

HOW JOB DESIGN MATTERS
N I C O L A I J . F O S S , D A N A B . M I N B A E VA ,
T O R B E N P E D E R S E N , A N D M I A R E I N H O LT
Job design is one of the most frequently researched practices in the Human
Resource Management (HRM) literature, and knowledge sharing has become
an important and heavily researched managerial practice. The links between
these practices, however, have received little attention in the literature. We
argue that job design matters to knowledge sharing for motivational reasons.
Specifically, jobs contain characteristics that stimulate different kinds of mo-
tivation toward knowledge sharing, which have different effects on individual
knowledge sharing behavior. We develop six hypotheses that unfold these
ideas and test them on the basis of individual-level data collected within a
single firm. The hypotheses are tested in a LISREL model that confirms that
job characteristics, such as autonomy, task identity, and feedback, determine
different motivations to share knowledge, which in turn predict employees’
knowledge sharing behaviors. © 2009 Wiley Periodicals, Inc.
Keywords: job design, job characteristics, types of motivation, knowledge

sharing
Introduction therefore be an important design variable for

firms that want to benefit from employees’
ob design has been one of the most sharing of relevant knowledge. Such sharing
J frequently researched Human Re-

source Management (HRM) prac-
tices. Among other things, the
importance of job design derives
from the impact it may have on employee
motivation (Lawler, Hackman, & Kaufman,
1973). We argue that job design may impact
may be particularly pertinent when there is a
risk that highly knowledgeable employees
may leave the organization or when there
are high costs of retaining such talent—
pressing problems for many consulting, ac-
counting, and professional services firms
that knowledge sharing may alleviate.
specific employee motivations to share The idea that job design can influence
knowledge. For this reason, job design is an pooling employee knowledge is not new.
antecedent of actual knowledge-sharing Job-sharing systems require knowledge over-
behaviors in organizations. Job design may lap, and firms may adopt self-managing teams
Correspondence to: Mia Reinholt, Center for Strategic Management and Globalization, Copenhagen Business
School, Porcelainshaven 24, 1st floor, 2000 Frederiksberg, Denmark, Phone: +45 3815 3392, Fax: +45 3815 3035,
E-mail: mr.smg@cbs.dk
Human Resource Management, November–December 2009, Vol. 48, No. 6, Pp. 871– 893
© 2009 Wiley Periodicals, Inc.
Published online in Wiley InterScience (www.interscience.wiley.com).
DOI: 10.1002/hrm.20320
872 HUMAN RESOURCE MANAGEMENT, NOVEMBER–DECEMBER 2009
to increase the overlap of employees’ knowl- because knowledge sharing takes place among
edge. Organization theorists have long main- the organization’s individual employees (cf.
tained that such outcomes of job design as Felin & Foss, 2006; Felin & Hesterly, 2007).
specialization may influence employee knowl- The following section introduces our
edge overlap and communication. Such rea- theoretical framework building blocks. We
soning, however, has usually placed cognitive draw on social psychology research on moti-
factors center stage. For example, heavy spe- vation (e.g., Deci & Ryan, 1985; Gagné & Deci,
cialization decreases the cognitive ability to 2005). Overall, we hypothesize a causal chain
absorb knowledge and may thus impede leading from (1) adopting certain kinds of job
knowledge sharing (e.g., Aoki, 1986; Lawrence design that (2) result in particular job charac-
& Lorsch, 1967). In contrast, we teristics that (3) impact employee motivation
argue that job design matters to to share knowledge, which in turn (4) affects
Management can knowledge sharing for fundamen- employees’ knowledge-sharing behaviors
tally motivational reasons. As has (sending and receiving knowledge). We put
design jobs to
been known at least since Hack- forward six hypotheses and test them on the
influence variables man and Oldham (1975; 1976), basis of individual-level data collected within
management can design jobs to a single firm. Conducting a one-site survey
such as autonomy, influence variables such as auton- allowed us to focus on individual-level issues,
omy, task identity, and the degree since many other factors, such as firm-, indus-
task identity, and the
of feedback the employee receives. try-, and country-specific factors, are kept
degree of feedback These job characteristics impact constant by design (cf. Siggelkow, 2007).
employee motivation to share
the employee knowledge, albeit in different ways,
Theoretical Framework and
and eventually affect knowledge-
receives. These
sharing behavior. Managers who
Constructs
job characteristics wish to design their organization
Knowledge Sharing
to pool employee knowledge bet-
impact employee ter should take these motivational Knowledge sharing is an important part of
effects into account. building knowledge-based competitive ad-
motivation to share Our argument links to the re- vantage (Argote & Ingram, 2000; Cohen &
knowledge, albeit in cent strategic HRM research stream Levinthal, 1990; Kogut & Zander, 1992).
on knowledge-based competitive Knowledge sharing can be studied and man-
different ways, and advantage (e.g., Kang, Morris, & aged at organizational, group, and individ-
Snell, 2007; Lepak & Snell, 1999, ual levels of analysis (Jackson, Chuang,
eventually affect 2002). A growing body of empiri- Harden, & Jiang, 2006). The premise of the
cal work (e.g., Cabrera, Collins, & present research, however, is that organiza-
knowledge-sharing
Salgado, 2006; Laursen & Foss, tional and group knowledge sharing are
behavior. 2003; Minbaeva, Pedersen, Björk- always ultimately rooted in individual be-
man, Fey, & Park, 2003) discusses haviors and their drivers (here, individual
HRM practices as means to im- motivation to share knowledge). More
prove firm-level knowledge sharing and cre- broadly, arguments that posit links between
ation. Yet, the literature fails to “adequately organizational variables, such as HRM prac-
address how firms can manage the knowledge tices, and organizational outcomes, such as
flows across different employee cohorts” (Kang organizational-level knowledge sharing,
et al., 2007, p. 243; italics in original). We must refer to individual-level mechanisms—
capture this by focusing on the impact of job that is, individuals’ motivations, cognition,
design on employees’ motivation to share and behaviors, and the interaction among
knowledge. We thereby provide an individual- those individuals (Coleman, 1990; Elster,
level perspective that is complementary to the 1989; Felin & Foss, 2006).
organizational-level emphasis in the strategic Knowledge sharing often involves mutual
HRM literature. This approach is warranted exchanges among individuals, including
Human Resource Management DOI: 10.1002/hrm

ENCOURAGING KNOWLEDGE SHARING AMONG EMPLOYEES 873
sending and receiving knowledge. It is a rela- and “motivation to make a prosocial differ-
tional act based on a sender-receiver relation- ence” (Grant, 2007). Although such recent
ship that incorporates communicating one’s contributions are clearly valuable extensions
knowledge to others as well as receiving oth- of Hackman and Oldham’s (1976) job charac-
ers’ knowledge (Hooff & de Leeuw van teristics theory, the three psychological states
Weenen, 2004). Knowledge, however, is often introduced in the original theory may en-
highly personal, not easily expressed, and compass such aspects. As developed by Hack-
thus difficult to share with others (Kogut & man and Oldham (1976) and Sims, Szilagyi,
Zander, 1992; Szulanski, 1996). Moreover, we and Keller (1976), job characteristics theory
cannot take employee motivation to share states that three groups of core job character-
knowledge for granted (Cabrera & Cabrera, istics activate the three critical psychological
2002; Osterloh & Frey, 2000). Organizational states mentioned previously. Accordingly, we
instruments can nevertheless be deployed to label the three job characteristics that corre-
foster knowledge sharing motivation and spond to these three psychological states
thus positively influence knowledge sharing. “autonomy,” “task identity,” and
While researchers have given considerable at- “feedback” (Fried, Oldham, &
tention to reward schemes (e.g., Beugelsdijk, Cummings, 2001), respectively. Knowledge sharing
2008; Minbaeva et al., 2003; Minbaeva, 2005), Autonomy concerns whether
is an important
they have seldom, if ever, explicitly consid- the job gives the employee the op-
ered job design as an antecedent of knowl- portunity to decide when and part of building
edge sharing. how to carry out specific tasks. In
other words, autonomy is “the knowledge-based
degree to which the job provides
Job Design and Job Characteristics competitive
substantial freedom, indepen-
Job design is a fundamental HRM activity. It dence, and discretion to the indi- advantage.
refers to deciding on the actual job struc- vidual in scheduling the work and
ture—that is, identifying the relevant tasks in determining the procedures to
and activities and allocating them across em- be used in carrying it out” (Hackman & Old-
ployees in a way that allows the organization ham, 1976, p. 258). Many studies have estab-
to reap benefits from specialization, as well as lished the importance of job autonomy by
bundling job tasks to take into account pos- finding positive relations between job au-
sible synergies between tasks. Traditionally, tonomy and, for example, the proclivity to
job design has focused on the job itself rather act proactively (Parker, Wall, & Jackson,
than on the specific individuals who are to 1997), personal initiative (Frese, Kring, Soose,
assume the job. Yet, researchers have long & Zempel, 1996), and felt responsibility
known that job design has motivational con- (Fuller et al., 2006; Parker et al., 1997). More-
sequences (e.g., Lawler et al., 1973). over, a high degree of autonomy may allow
Jobs possess certain characteristics that the employee to free up time for learning and
have psychological implications. The three development (Latham & Pinder, 2005, p.
critical psychological states that the relevant 493).
literature has focused on are (1) the experi- Task identity is the degree to which the
enced meaningfulness of the work, (2) the job gives the employee the opportunity to
experienced responsibility for work outcomes, undertake job tasks from beginning to end.
and (3) knowledge of the actual results of That is, task identity relates to whether an
one’s own work efforts (e.g., Parker & Wall, employee completes a whole, identifiable
1998). Recently, researchers have extended piece of work (Hackman & Oldham, 1976).
this focus with constructs such as “felt re- Identifying with one’s job tasks is important
sponsibility for constructive change” (Fuller, for how meaningful the employee perceives
Marler, & Hester, 2006), “perceived social the job to be, which has considerable impli-
impact” (Grant, 2008a), “trust by others and cations for an employee’s motivation in a
self” (Clegg & Spencer, 2007; Langfred, 2007), given job (Gagné & Deci, 2005).

Feedback is the degree to which the em- thus, the reason underlying the behavior is
ployee receives direct and clear information not inherent in the behavior itself but rather
about his or her performance as the person is instrumental in obtaining separate out-
carries out the job tasks. Hackman and Old- comes. Typical external motivators include
ham (1975) distinguish between two types of external rewards, such as money and praise,
feedback: one that is a characteristic of the as well as avoidance of punishment. Typically,
job itself and one that stems from other other individuals such as managers, col-
agents such as managers (e.g., leagues, and customers administer such con-
through practices such as evalua- tingencies. Individuals thus feel pressured
Intrinsic motivation
tion and recognition schemes or from the outside when externally motivated.
to engage in more informal verbal feedback). In contrast, intrinsic motivation involves
Strictly speaking, the latter type of doing an activity because it is in accord with
knowledge sharing feedback is not conventionally the individual’s intrinsic interest and per-
seen as a job characteristic but sonal values (Ryan & Deci, 2000). The em-
implies employees
rather as an HRM practice. There ployee derives enjoyment from conducting
find the activity itself is a case to be made, however, for the behavior per se (Deci, 1971; Gagné &
the notion that both types of Deci, 2005; Lepper, Greene, & Nisbett, 1973).
interesting, enjoying, feedback are important to the em- An intrinsically motivated individual is
ployees’ perception of their job mainly absorbed in the process of doing
and stimulating. characteristics. Receiving feedback an activity, whereas an externally motivated
on one’s performance is a critical individual is concerned with the external
element of feeling competent in the job and outcome attained from doing the activity.
is thus a strong predictor of motivation and Intrinsic motivation therefore implies that
performance (Deci, Koestner, & Ryan, 1999). the individual is free from pressure and ten-
sion when engaging in a particular behavior
(Deci & Ryan, 1985).
Types of Motivation
Deci and Ryan (1985, 2000) argue that we
Important theories of work motivation ap- need an even more fine-grained motivation
proach it as a unitary concept that solely var- typology to understand human behavior and
ies in strength but not in kind (e.g., Locke & its consequences more fully. An individual
Latham, 2002; Vroom, 1964). Different types can internalize external demands such that
of motivations, however, may lead to differ- behavior is self-regulated yet not intrinsically
ent performance qualities (Deci & Ryan, motivated. Consequently, Deci and Ryan sug-
2000; Vansteenkiste, Simons, Lens, Sheldon, gest additional motivation types, including
& Deci, 2004). Moreover, “differentiating mo- introjection, which occurs when an individual
tivation and goals provides an integrated “takes in” an external regulation but does not
means of relating characteristics of tasks and accept it as his or her own (Deci & Ryan,
interpersonal environments, as well as indi- 1985). This implies that the behavior is no
vidual differences, to types of performance longer contingent on others’ external rewards
and well-being” (Gagné & Deci, 2005, p. 341). and punishments. Instead, the individual
That is, motivation management can better monitors and administers sanctions and re-
be tailored to the specific needs of a particular wards to himself or herself (Deci & Ryan,
job when different types of motivation are 1985). An important hallmark of introjected
taken into account. Related to this reasoning, motivation is to promote feelings of worth
some work motivation theories distinguish (Ryan, 1995; Ryan, Koestner, & Deci, 1991).
between different types of motivation— Introjected motivation is in accordance with
notably, intrinsic and external motivation the “ought self-regulation” of regulatory
(e.g., Frey, 1997; Osterloh & Frey, 2000). focus theory (Higgins, 1997). That is, the in-
External motivation means that an indi- dividual is not acting on verbalized expecta-
vidual engages in an activity to attain a posi- tions and demands but rather on how the
tive or to avoid a negative external outcome; person believes others want him or her to

behave. In this sense, the individual feels likely to matter for knowledge sharing (Ca-
controlled by an external source. brera et al., 2006; Lin, 2007; Osterloh & Frey,
2000). Empirically, researchers have consis-
Hypotheses Development tently shown that the impact of motivational
factors such as self-efficacy, development, and
We are interested in understanding how dif- enjoyment often associated with intrinsic
ferent aspects of job design, through their motivation enhance knowledge sharing (e.g.,
impact on job characteristics, foster certain Bock, Zmud, Kim, & Lee, 2005; Burgess, 2005;
types of intrinsic, introjected, and external Cabrera et al., 2006; Lin, 2007; Quigley, Tes-
motivation, as well as how these types of luk, Locke, & Bartol, 2007). We therefore
motivation influence an employee’s knowl- expect that intrinsic motivation is positively
edge-sharing behavior. We develop our related to knowledge sharing:
hypothesesin the following; Figure 1 summa-
rizes the theoretical model. Hypothesis 1: The more intrinsically motivated
employees are to share knowledge, the more knowl-
edge they will (a) receive from and (b) send to
The Motivation to Share Knowledge
colleagues.
Knowledge Sharing and Intrinsic
Motivation Knowledge Sharing and Introjected
Motivation
Intrinsic motivation to engage in knowledge
sharing implies employees find the activity Employees motivated by introjection are
itself interesting, enjoying, and stimulating. concerned with maintaining and enhancing
In a broader perspective, social psychology feelings of worth in their social groups
research argues that intrinsically motivated (Koestner & Losier, 2002; Ryan & Deci,
individuals are proactive and get involved in 2000). In an organizational context, doing a
activities to promote their own personal good job that is socially acceptable and that
growth (Deci & Ryan, 2000). Many empirical other organizational members view as ap-
studies show intrinsic motivation promotes propriate becomes a prime mover of the
highly valued behavioral outcomes, such as employee’s actions. We predict that this mo-
creativity (Amabile, 1993), quality (Kruglan- tivation type is positively related to knowl-
ski, Friedman, & Zeevi, 1971), and learning edge sharing. First, if an employee believes
(Vallerand & Bissonnette, 1992; Vansteen- there is an expectation regarding knowledge
kiste et al., 2004). sharing within the organization, he or she
It is reasonable to expect that intrinsic will make an effort to comply with that ex-
motivation will have the same positive effects pectation to maintain feelings of worth.
on knowledge sharing as it has on other This will lead to high involvement in knowl-
learning activities. In fact, several scholars edge sharing (cf. Bock et al., 2005). Second,
argue that intrinsic motivation is particularly when an employee wants to fit in and gain
H4(+)
Autonomy Intrinsic motivation H1(+)
H5(+) H2(+)
Task identity Introjected motivation Knowledge sharing
H6(+)
Feedback External motivation H3(+)
FIGURE 1. Theoretical Model

acceptance in the organization, it becomes In addition to the valued outcomes the em-
important for him or her to solve tasks in ployee obtains from engaging in knowledge
accordance with how managers and col- sharing itself, he or she will reap the benefits
leagues expect them to be solved. The em- from others’ knowledge when doing his or
ployee will be more inclined to rely on her own work assignments, which may lead
others’ input and knowledge of how to solve to more of the valued external outcome. In a
the task correctly and thereby minimize his workplace setting where employees repeat-
or her risk of failure. Furthermore, the em- edly interact, we would expect reciprocal
ployee will send knowledge to colleagues to behavior to arise. We thus predict that em-
gain acceptance in the organization. This ployees who receive knowledge from others
suggests the following hypothesis: and are externally motivated to seek out such
knowledge will send knowledge to colleagues
Hypothesis 2: The more introjectedly in an ongoing knowledge-trading game. In
If an employee motivated employees are to share sum, we hypothesize that:
knowledge, the more knowledge they
believes there is will (a) receive from and (b) send to Hypothesis 3: The more externally motivated
colleagues. employees are to share knowledge, the more
an expectation
knowledge they will (a) receive from and (b)
regarding knowledge Knowledge Sharing and External send to colleagues.
Motivation
sharing within the Job Autonomy
Expectancy theory (Vroom, 1964)
organization, he or
and expected utility theory assert When a specific job is designed to provide the
she will make an that an individual’s expectations individual with autonomy in planning and
of gains and losses based on sub- conducting the job, his or her sense of respon-
effort to comply with jective probability estimates are sibility for work-related outcomes is enhanced
the basis for his or her motivation (Fuller et al., 2006; Hackman & Oldham,
that expectation to to engage in a certain behavior. 1975). At a general level job, therefore, au-
maintain feelings of Most knowledge-sharing literature tonomy (or freedom in the job) is an impor-
investigating the role of motiva- tant mechanism impacting employees’
worth. This will lead tion is arguably based on such motivation and thus their performance. The
reasoning (e.g., Cabrera & Cabrera, social psychology literature broadly supports
to high involvement 2002), in which motivating knowl- the argument that the experienced level of
edge sharing is a matter of restruc- autonomy in the environment—including,
in knowledge
turing the payoff function so the but not limited to, the autonomy the job itself
sharing. employee finds knowledge sharing provides— influences motivation (Gagné &
beneficial. Other types of external Deci, 2005). The main argument in this litera-
motivational factors include for- ture is that all individuals have three basic
mal recognition and feedback, which several psychological needs—for competence, auton-
scholars consistently argue have a strong posi- omy, and social relatedness—and that satisfy-
tive effect on knowledge sharing (e.g., Cabrera ing these needs will lead to highly motivated
& Cabrera, 2005; O’Dell & Grayson, 1998). employees. In particular, the need for auton-
The main characteristic of an employee omy is important to the extent that there is a
externally motivated toward knowledge shar- need to maintain intrinsic motivation (Deci &
ing is that some valued external contingency Ryan, 2000). Empirical studies have widely
drives his or her involvement in knowledge established the importance of autonomy for
sharing. Irrespective of the particular external intrinsic motivation and performance. Deci
motivational factor in question, we therefore and Ryan (2000) found that an autonomous
argue that external motivation is positively work environment is positively related to in-
related to knowledge sharing when the em- trinsic motivation, higher performance, and
ployee values the expected external outcome. employee engagement in work. Gagné (2003)

found that autonomy was strongly related to that he or she no longer needs external con-
students’ engagement in voluntary work as tingencies to prompt the desired behavior
well as other prosocial activities. This indi- (Gagné & Deci, 2005). In other words, the
cates a positive relation between autonomy employee will tend to internalize the value of
and intrinsic motivation toward knowledge tasks to the extent that he or she thinks the
sharing. By designing jobs to enhance the em- tasks make sense and there is a meaningful
ployees’ experience of autonomy, managers rationale behind conducting them. Since
will thus positively impact their intrinsic mo- following a task through all its stages makes
tivation to engage in knowledge sharing: the job more meaningful to the employee
(Hackman & Oldham, 1976), he
Hypothesis 4: The more autonomy the job pro- or she will initiate such an inter-
vides, the more the employee will tend to be in- nalization process. Introjected mo- By designing
trinsically motivated to share knowledge. tivation is thus the result. We
jobs to enhance
therefore hypothesize that:
Task Identity the employees’
Hypothesis 5: The greater the degree of
In addition to job autonomy, Hackman and task identity the job provides, the more experience of
Oldham (1975) argue that task identity is an the employee will tend to be introject-
autonomy, managers
important aspect of the job that may lead to edly motivated to share knowledge.
high levels of what they call “internal motiva- will thus positively
tion.” They argue that once employees begin Feedback
to understand their tasks as a whole and iden- impact their intrinsic
tifiable piece of work, they will perceive the Feedback on the job is another
motivation to engage
job as more meaningful. Hackman and Old- mechanism researchers argue to
ham (1975), however, do not distinguish be- be important to an employee’s in knowledge
tween different types of internal motivation motivation, satisfaction, and per-
(e.g., intrinsic and introjected motivation). formance. Hackman and Oldham sharing.
They do not explore how different aspects of (1975) argue that an employee
their job design framework promote different who receives feedback as a natural
types of internal motivation and the possible part of the job will tend to experience
performance differences they may cause. We the positive outcomes mentioned earlier.
address this issue by including both intrinsic Job design and characteristics theory also
and introjected motivation. emphasizes the importance of other types of
Whereas job autonomy is associated with feedback such as those from the employee’s
the freedom to plan and carry out the job in manager and colleagues (Hackman & Old-
ways the employee finds most suitable, task ham, 1975; Sims et al., 1976).
identity refers to the employee’s following a There is a natural link between external
task through all its stages so he or she, for feedback and external motivation: When an
instance, “provides a complete unit of ser- employee is motivated by feedback in the
vice” (Hackman & Oldham, 1976, p. 257) form of, for example, evaluations and recog-
instead of just part of it. Task identity con- nition schemes, he or she will behave in cer-
cerns identifying with a task or a job others tain ways to obtain attractive evaluations or
define as well as following procedures others recognition. External contingencies adminis-
formulate. Designing jobs in accordance with tered by other individuals thus regulate em-
this job design dimension is therefore likely ployee behavior. Note that researchers have
to involve certain expectations regarding found unexpected positive feedback to en-
how the task should be solved as well as its hance intrinsic motivation, whereas more
performance outcomes. controlling types of feedback such as antici-
When an employee identifies with his or pated evaluation and positive feedback deliv-
her tasks, however, the person may internal- ered in a controlling manner are shown to
ize external demands and expectations so have negative effects on intrinsic motivation

(e.g., Deci et al., 1999; Harackiewicz, Mander- ing (cf. Siggelkow, 2007). This is a major ad-
link, & Sansone, 1984). A shift in the vantage compared to questionnaires designed
individual’s perceived locus of causality often to target a large number of firms but only one
explains this negative effect—which means or a few respondents per organization.
that the person shifts from feeling MAN Diesel is the world’s leading pro-
like the initiator of his or her own vider of large-bore diesel engines for marine
Feedback behavior to behaving for external and power plant applications. It generates
reasons (in this case, to obtain a revenue through license royalties. Secondary
mechanisms such
positive evaluation and recogni- business areas include resale of engines, com-
as recognition tion). In other words, external ponent sales, and introducing new features
rather than intrinsic motivation is for engines already in operation. MAN Diesel
and performance stimulated. It therefore seems rea- employs more than 6,400 staff members, pri-
sonable to expect that formal types marily in Germany, Denmark, France, the
evaluations tied to
of feedback are positively related United Kingdom, the Czech Republic, and
knowledge-sharing to external motivation. China. The Danish subsidiary has designed
Related to this, several scholars and produced two-stroke engines for more
performance theorize that feedback mecha- than 100 years. Engineers and designers lo-
nisms such as recognition and cated in Copenhagen undertake all research
may signal that
performance evaluations tied to and development (R&D), design, and testing,
knowledge sharing knowledge-sharing performance while the majority of the production (92%)
may signal that knowledge shar- takes place at the licensees in Japan, Korea,
is important to, ing is important to, and valued by, and China.
the organization and thereby posi- MAN Diesel in Copenhagen is in many
and valued by, the tively impact motivation to en- respects a typical engineering company, em-
organization and gage in knowledge sharing (e.g., ploying mainly machine engineers and other
Cabrera & Cabrera, 2005; O’Dell & engineers (e.g., in electronics and information
thereby positively Grayson, 1998). Furthermore, a technology) with an academic degree. The
study of determinants of knowl- organizational structure is hierarchical and
impact motivation edge sharing finds evidence for strongly departmentalized, with clear lines of
the importance of feelings of self- responsibility flowing from the top to the bot-
to engage in
efficacy, which may further indi- tom. The business model, however, requires
knowledge sharing. cate that, for example, recognition concerted efforts to navigate and nurture
schemes that strengthen the em- various kinds of relationships with stakehold-
ployee’s feeling of self-efficacy can ers and customers, such as shipyards, ship
have positive effects on knowledge sharing owners, classification societies, authorities,
(Cabrera et al., 2006). We thus expect: and suppliers. The ability to share knowledge
and facilitate innovations within and across
Hypothesis 6: The more feedback in the form of departments is therefore a key concern to
formal evaluations and recognition schemes the management.
job provides, the more the employee will be exter-
nally motivated to share knowledge. Research Instrument
Methods We collected the data using an administered

questionnaire based on a focused literature
We collected all data used in the analysis review. We pretested the questionnaire with
based on one-site sampling (Tsai & Ghoshal, academicians and managers to ensure that
1998) at the Copenhagen site of the German individual items and the overall format were
multinational company MAN Diesel. This easily understood. Further, we tested the
implies that our research design controls for a questionnaire with the representatives of
number of broad contextual factors known to MAN Diesel to increase the clarity of the
influence intraorganizational knowledge shar- questions and avoid interpretation errors.

We used self-reported or perceptual mea- with data from two follow-up meetings that
sures for operationalizing all variables in the each lasted two hours with executive officers
questionnaire. Self-reported measures are from MAN Diesel. The discussion of the re-
particularly useful in providing a picture of sults took place in two stages. First, we pre-
how people perceive and feel about their job- sented the results and pointed out some
related behavior (Spector, 1994). Researchers preliminary findings and reflections on the
recommend perceptual measures for studies results. We discussed those findings; in the
of human behavior in general (Howard, majority of the cases, the firm executives of-
1994) and studies on motivation in particu- fered alternative explanations. In return,
lar. Most other studies on intraorganizational they pointed to those issues that, in their
knowledge processes also rely on perceptual opinion, were the most interesting and rele-
measures (e.g., Gupta & Govindarajan, 2000; vant for MAN Diesel. Such design allowed
Lyles & Salk, 1996; Simonin, 1999; Szulanski, our ideas to “be hatched, tested, and
1996). Finally, using the perceptual measures [dis]confirmed in a relatively short period of
allowed us to capture the implemented job time” (Chatman & Flynn, 2005, p. 439). It
design practices, or practices in use that also allowed us to contextualize our findings
were “perceived and interpreted subjectively to the extent possible (cf. Eisenhardt &
by each employee” (Wright & Niishi, 2006, Graebner, 2007, p. 25).
p. 11).
The firm requested a Web-based version Measures
of the questionnaire. A firm representative,
who mediated collection of the question- In the following sections, we describe the
naires, distributed the invitation containing operationalization of the constructs and we
the link to the Web-based survey via the then evaluate the different forms of validity.
firm’s internal e-mail system. To reduce pos- See Table II for the exact wording of ques-
sible social desirability bias, we followed Tsai tions forming the items.
and Ghoshal (1998), explaining in the ques-
tionnaire introduction that the software pre-
Knowledge Sharing
vented identification of individuals and that
the data would be collected using an external Knowledge sharing includes the respondent’s
server independent of the firm. both receiving knowledge and sending it.
Together with the firm representative, we Gupta and Govindarajan (2000) pointed out
selected those departments for the survey that the importance of distinguishing between
are mostly involved in knowledge sharing receiving and sending knowledge. Daven-
across individuals and departments (i.e., Engi- port and Prusak (1998) similarly argue that
neering, R&D, Sales & Marketing, Technical knowledge sharing involves two actions: the
Service, and Purchasing). We submitted the recipient’s transmission and absorption/use
questionnaire to all individuals in those de- of the knowledge. Accordingly, for the receiv-
partments. Of the 505 invitations sent out for ing of knowledge, we asked individual respon-
participation in the survey, 263 question- dents to indicate the extent to which they
naires were filled in. Because of missing values have gained and used knowledge from col-
for some items, however, we were only able to leagues in their own department. Similarly,
use 186 responses in the data analysis. This for the sending of knowledge, we asked respon-
equals a usable response rate of 33%. Table I dents about the extent to which colleagues
presents a description of the respondents. in their own department have gained and
After consultation with the firm representa- used the respondent’s knowledge. We ad-
tive, we regarded as representative the distri- opted these measures from Minbaeva et al.
bution of the survey responses. (2003) but modified them to the individual
Although we tested the hypotheses using level. The four questions used a 7-point scale
the survey data, to make our conclusions from 1 (little or no extent) to 7 (very large
more robust we triangulated the survey data extent).

The Motivation to Share Knowledge jected, and external motivation. We adopted

the scales from the Self-Regulation Question-
As pointed out earlier, we approach motiva- naires, which assess different types of motiva-
tion as a multidimensional construct. Follow- tion (Ryan & Connell, 1989). Using a 7-point
ing Deci and Ryan (1985) and Ryan and Deci scale ranging from “strongly disagree” to
(2000), we distinguish among intrinsic, intro- “strongly agree,” we applied three items/
TABLE I Response Distribution

Responses
Distribution by Gender
Male 159
Female 26
Nonresponse 1
Distribution by Age
25–34 40
35–44 68
45–54 44
55–64 31
Nonresponse 3
Distribution by Education
High School or Below 10
Middle-range Training 31
Bachelor’s Degree 77
Master’s Degree 34
PhD 8
Other 23
Nonresponse 3
Distribution by Years at MAN Diesel
<5 42
5–9 38
10–14 36
15–19 30
20–29 22
> 29 15
Nonresponse 3
Distribution by Department
Engineering 95
R&D 47
Sales & Marketing 11
Technical Service 16
Purchasing 12
Nonresponse 5
Stilling
Top management 24
Technical (Engineer + Project Manager + Other Technical) 141
Administration (Administrative + Other + Sales) 18
Nonresponse 3
Total 186

TABLE II Constructs and Items*

Average
Constructs and Items Variance
Factor Construct Extracted
Loading t-value R 2 Value Reliability (AVE)
Receiving of Knowledge
To what extent have … 0.88 0.79
… you received knowledge from colleagues 0.85 12.41 0.73
in your own department?
… you used knowledge from colleagues in your 0.93 13.68 0.86
own department?
Sending of Knowledge
To what extent have colleagues … 0.93 0.87
… in your own department received knowledge 0.99 16.20 0.99
from you?
… in your own department used knowledge from you? 0.87 13.51 0.76
Intrinsic Motivation
Why do you share knowledge with others? 0.76 0.51
I think it is an important part of my job. 0.69 8.80 0.47
I find it personally satisfying. 0.68 6.27 0.46
I like sharing knowledge. 0.78 9.87 0.60
Introjected Motivation
I share knowledge because … 0.87 0.69
… I feel proud of myself. 0.72 10.86 0.52
… I want my superior to think I am a good employee. 0.90 14.89 0.80
… I want my colleagues to think I am competent. 0.87 14.33 0.76
External Motivation
Why do you share knowledge with others? 0.87 0.64
I want my supervisor(s) to praise me. 0.97 17.57 0.93
I want my colleagues to praise me. 0.94 16.76 0.88
I might get a reward. 0.66 9.90 0.43
It may help me get promoted. 0.58 8.44 0.34
Autonomy
To what extent is your job characterized by the following? 0.76 0.51
The freedom to carry out my job the way I want to. 0.72 9.94 0.52
The opportunity for independent initiative. 0.83 11.61 0.69
High level of variety in my job. 0.57 7.47 0.33
Task Identity
To what extent is your job characterized by the following? 0.76 0.52
The opportunity to complete work that I started. 0.74 7.79 0.54
The opportunity to do a job from the beginning to 0.75 8.51 0.56
the end.
The opportunity to do my job independently of others. 0.67 5.64 0.45
Feedback
To what extent are you included in the following? 0.77 0.53
Formal acknowledgment. 0.70 9.59 0.50
Performance evaluation. 0.82 11.45 0.67
Feedback from my superior on my job performance. 0.66 9.01 0.44
*All variables are measured on a 7-point Likert scale.

questions to capture intrinsic and introjected Non-Normed Index (NNFI) = 0.87, the single
motivation, while we based the external mo- factor model did not fit the data well. Further-
tivation construct on four items. more, the improved fit of the alternative
and more complex models (also reported in
Table IV) was statistically significant. While
Job Characteristics
these tests do not eliminate the threat of com-
We adopted measures of job characteristics mon method variance, they provide evidence
from Sims et al. (1976), who used Hackman that common method bias does not purely
and Oldham’s (1975) contribution to de- drive interitem correlations.
velop an improved instrument for job char- The correlation matrix in Table III pro-
acteristics (the Job Characteristic Inventory) vides further evidence that the data do not
and provided evidence for the instrument’s suffer from common method bias. In fact,
reliability and validity. For both autonomy the correlation matrix shows that the corre-
and task identity, we used three items on a lations, in general, are much higher inside
7-point scale (from “very little extent” to the constructs (as expected) than all other
“very large extent”) to measure these vari- coefficients. For some constructs like the re-
ables. We also captured the variable feedback ceiving and sending of knowledge and intro-
by three items. We measured two of the prac- jected and external motivation, however, the
tices—“formal acknowledgement” and “per- items still have relatively high across-
formance evaluation”—on a 7-point scale construct correlations (in the range of 0.35
(from “very little extent” to “very large ex- to 0.50), which call for tests of alternative
tent”), while we measured the third item on specifications of the model.
a 7-point scale ranging from “not at all or
very little” to “very much.” Validity and Reliability of Measures
With respect to common method bias, we
placed the performance variables after the We tested the hypotheses in a LISREL model
independent variables in the survey to dimin- that allows for simultaneous formation of
ish, if not avoid, the effects of consistency underlying constructs (the measurement
artifacts (Salancik & Pfeffer, 1977). In addi- model) and test of structural relationships
tion, we performed a number of statistical among these constructs (the structural model).
tests to detect potential common method To ascertain whether the constructs are inter-
bias. First, we conducted a Harman’s one-fac- nally coherent, we report several tests of con-
tor test on the items included in our model. vergent validity in Table II, which is based on
Here we found multiple factors (seven factors the saturated measurement model in which
with an eigenvalue > 1), and the first two fac- all interfactor correlations are specified (Jo-
tors accounted for only 21% and 17% of the reskog & Sorbom, 1993). First, the strengths
variance, respectively (Podsakoff & Organ, of linearity are relatively strong with R2 values
1986). Second, we conducted the stronger test of 0.33 or above, which is clearly above the
of common method bias—the “single factor usual threshold of 0.20 (Hair, Anderson, Ta-
procedure”—based on confirmatory factor tham, & Black, 1995). From Table II we can
analyses Podsakoff, MacKenzie, and Podsakoff also conclude that the t-values for all items
(2003) recommend for this kind of study. We are highly significant (they are all above 5.64)
examined the fit of the single factor model in and that their (standardized) factor loadings
which all items loaded on one factor. The un- are strong (all above 0.57). Second, we calcu-
derlying logic is that if method variance is lated the correspondence between the items
largely responsible for the covariation among and their constructs (i.e., the construct reli-
the constructs, a confirmatory factor analysis ability) for each construct as the share of
should indicate that a single factor model fits common variance the construct explains. All
the data. Goodness-of-fit statistics for the eight constructs are above the recommended
single factor model are listed in Table IV; with threshold of 0.70 (see column 5 in Table II)
the Goodness-of-Fit Index (GFI) = 0.85 and (Gerbing & Anderson, 1988). The construct

TABLE III Correlation Matrix (N = 186)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
1. Received knowledge 0.89
from colleagues
2. Used knowledge from 0.79 0.89
colleagues
3. Colleagues received 0.39 0.38 0.93
knowledge from you
4. Colleagues used 0.33 0.35 0.86 0.93
knowledge from you?
5. Important part of my job 0.18 0.29 0.34 0.32 0.72
6. Find it personally 0.15 0.20 0.19 0.16 0.52 0.72

satisfying
7. Like sharing knowledge 0.25 0.33 0.34 0.31 0.53 0.60 0.72
8. Feel proud of myself -0.01 0.01 0.02 0.01 -0.01 0.22 0.09 0.83
9. Want my superior to think 0.02 0.03 -0.07 -0.05 -0.06 0.17 0.04 0.63 0.83
I am a good employee
10. Want my colleagues to 0.01 -0.02 -0.01 0.03 -0.06 0.28 0.08 0.63 0.79 0.83
think I am competent
11. Want my supervisor(s) to 0.03 -0.03 0.01 0.04 -0.12 0.14 0.01 0.50 0.46 0.49 0.80
praise me
12. Want my colleagues to 0.06 -0.0 0.01 0.05 -0.1 0.14 0.01 0.42 0.45 0.42 0.91 0.80
praise me
13. Might get a reward -0.02 -0.07 0.01 -0.02 -0.07 0.10 -0.02 0.38 0.43 0.37 0.63 0.59 0.80
14. May help me get -0.01 -0.06 -0.04 -0.01 -0.06 0.11 0.08 0.38 0.41 0.46 0.56 0.60 0.54 0.80
promoted
15. Freedom to carry out my 0.09 0.15 0.26 0.21 0.13 0.03 0.24 0.13 0.09 0.20 -0.03 0.01 -0.04 0.02 0.71
job the way I want
16. Opportunity for 0.20 0.27 0.32 0.26 0.12 0.02 0.24 0.16 0.08 0.14 0.01 0.01 -0.03 0.06 0.62 0.71
independent initiative
17. High level of variety 0.29 0.29 0.33 0.31 0.10 0.11 0.17 0.08 0.06 0.14 0.11 0.13 0.07 0.10 0.57 0.55 0.71
in my job
18. Opportunity to complete 0.12 0.15 0.01 0.03 -0.03 -0.04 0.11 0.04 0.08 0.10 -0.06 -0.11 -0.01 0.08 0.27 0.34 0.25 0.72
work that I started
19. Opportunity to do my job -0.07 -0.08 -0.04 -0.04 -0.09 -0.02 0.01 -0.04 -0.01 0.05 -0.09 -0.11 -0.01 -0.03 0.15 0.08 0.06 0.54 0.72
independently of thers
ENCOURAGING KNOWLEDGE SHARING AMONG EMPLOYEES
883
884
TABLE III Continued.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
20. Opportunity to do a job 0.03 0.07 0.05 0.08 -0.07 0.01 0.10 0.10 0.13 0.16 -0.04 -0.06 0.01 -0.01 0.18 0.17 0.16 0.57 0.61 0.72
from the beginning to
the end
21. Formal acknowledgment 0.20 0.18 0.17 0.17 0.12 0.21 0.13 0.08 0.08 0.11 0.21 0.17 0.21 0.10 0.18 0.25 0.28 0.19 0.01 0.10 0.73
22. Performance evaluation 0.32 0.31 0.15 0.16 0.12 0.16 0.14 0.17 0.11 0.17 0.19 0.20 0.28 0.10 0.11 0.20 0.27 0.14 0.06 0.11 0.57 0.73
23. Feedback from my 0.28 0.27 0.10 0.15 0.11 0.10 0.13 0.07 0.07 0.09 0.14 0.14 0.16 0.05 -0.01 0.11 0.25 0.05 0.03 0.09 0.56 0.54 0.73
superior on job
HUMAN RESOURCE MANAGEMENT, NOVEMBER–DECEMBER 2009
performance
Mean 5.78 5.96 5.53 5.44 6.16 5.80 5.99 4.23 4.20 4.74 3.45 3.63 3.10 3.20 5.92 6.06 4.97 5.37 5.25 5.35 3.41 3.26 3.41
Std. Dev. 1.21 1.11 1.22 1.25 0.84 1.16 0.90 1.65 1.75 1.57 1.56 1.61 1.39 1.60 1.24 1.20 1.52 1.33 1.42 1.41 1.64 1.64 1.71
All coefficients above 0.15 are significant at the 5% level.

reliability varies from 0.76 (three constructs) erable evidence for discriminant validity of
to 0.92 (the construct for sending of knowl- each of our eight constructs. In sum, there is
edge). Furthermore, Fornell and Larcker (1981) strong evidence for validity of our eight con-
proposed a statistic they termed Average structs. This is also reflected in the goodness-
Variance Extracted (AVE) as a measure of of-fit statistics for the whole measurement
convergent validity (i.e., the measure of the model that with GFI = 0.90, NNFI = 0.96, and
error-free variance of the set of items related RMSEA = 0.04 meets the requirements for ac-
to a construct). With regard to the AVE statis- cepting the model.
tics, the constructs are clearly robust, as all
constructs are above the recommended The Goodness-of-Fit of the
threshold of 0.50. Structural Model
We obtained several measures of discrimi-
nant validity from the data. One suggested test The second step in the analytical process is
of discriminant validity is the test of whether to form the structural model by specifying
the correlations and causal paths between the the causal relations in accordance with the
latent constructs are significantly different hypotheses. Through repeated iterations, a
from 1 (Fornell & Larcker, 1981). By con- LISREL analysis fine-tunes the model to ob-
structing 99.9% confidence intervals around tain a more coherent representation of the
the correlations and causal paths, we can con- empirical data. We thus generated a struc-
firm that none of them is close to 1. In addi- tural model that contains relationships in
tion, one can use the AVE statistics to assess accordance with the stipulated hypotheses.
discriminant validity. If the square root of In addition, we allowed the three types of
AVE is larger than the correlation with items motivation to correlate as we expect some
belonging to other constructs, this suggests level of correlation between them. We tested
that each construct has more internal (ex- single causal relations with t-values and fac-
tracted) variance than variance shared with tor loadings between the constructs in the
other constructs, which indicates the focal model. Goodness-of-fit indexes are critical
construct is indeed different from other con- for the evaluation of the entire model. Given
structs (i.e., discriminant validity). The diago- their complexity, however, there is no con-
nal of Table III shows the square root of sensus regarding the “best” index of overall
AVE-value for all constructs, and none of the fit for structural equations. Reporting multi-
correlation coefficients exceeds the values of ple indexes is thus encouraged (Bollen,
the square root of AVE. This provides consid- 1989).
TABLE IV Goodness-of-Fit Statistics for Three Competing Specifications of the Model

0 1 2 3 4
Single Measurement Each Job Design Dependent Variable Theoretical
Factor Model Variable Linked to All = Receiving + model
Model Three Motivation Sending of
Variables Knowledge
Chi-square (d.f.) 418.8 256.4 319.2 561.5 334.0
(214 d.f.) (202 d.f.) (204 d.f.) (214 d.f.) (218 d.f.)
GFI 0.85 0.90 0.88 0.82 0.90
GFI adjusted for d.f. 0.80 0.86 0.83 0.76 0.86
Parsimonious GFI 0.71 0.72 0.71 0.69 0.75
RMSEA 0.07 0.04 0.06 0.09 0.05
Comparative fit 0.89 0.97 0.94 0.82 0.94
index
NNFI 0.87 0.96 0.93 0.79 0.93
Parsimonious NFI 0.68 0.71 0.69 0.63 0.73

We assessed the structural model by dif- a bad fit (high chi-square values, etc.) and is
ferent goodness-of-fit measures, including clearly not an acceptable description of the
the chi-square value, the GFI, and the NNFI, data. We must therefore reject it. The model
which are measures of the distance between linking all job characteristics with all moti-
data and model—that is, nomological valid- vation variables performs better and is a bor-
ity (Joreskog & Sorbom, 1993). The theoreti- derline case in the sense that some measures
cal and hypothesized model (i.e., model 4 in such as RMSEA and NNFI are acceptable,
Table IV) has a chi-square value of 2[218] = while others such as GFI are not. The bottom
334.0 ( p = 0.01), while the GFI based on line is that the theoretical model is clearly
residuals obtains a value of 0.90, which rep- the most parsimonious model of the four
resents a good fit of the model to the data specifications listed in Table IV. With a parsi-
(Bollen, 1989). Finally, the Bentler-Bonett monious GFI of 0.75 and parsimonious NFI
NNFI represents the proportion of improve- of 0.73 (that adjust for the higher degrees
ment in fit relative to the null model, while of freedom in the theoretical model), the
controlling for model parsimony. The ob- theoretical model clearly obtains a better
tained value (NNFI = 0.93) represents an overall goodness-of-fit than the alternatives
equally good fit of the model to the data. In and therefore provides the best description
addition, the RMSEA is only 0.05 and there- of the data.
fore is below the suggested threshold of 0.08.
The conclusion thus based on the three mea- Results
sures GFI, NNFI, and RMSEA is that we ob-
tained a good fit of the proposed model to The three hypotheses that link job design and
the data. job characteristics to motivation (H4 through
Furthermore, we compared the theoreti- H6) are strongly supported (see Figure 2). We
cal model with a number of its alternative find that autonomy is positively (coefficient:
specifications, including the saturated mea- 0.39) and significantly (p < 0.01) related to the
surement model. Table IV provides the com- degree of intrinsic motivation, and that the
parable statistics for the alternative models. degree to which the job contains task identity
In addition to being compared to the satu- is positive and significantly associated with
rated measurement model, the theoretical introjected motivation (coefficient: 0.23 and
model is compared with two alternative p < 0.01). Feedback positively and signifi-
specifications of the model. In one alterna- cantly influences external motivation (coeffi-
tive specification (model 2 in Table IV), each cient: 0.20 and p < 0.01).
of the three variables of job characteristics is H1 through H3 concern the relationships
allowed to influence all three motivation between the three motivation variables and
variables (i.e., 3 x 3 relationships instead of the sharing of knowledge. We find that intrin-
only the three hypothesized relationships in sic motivation strongly and positively affects
H4 through H6). By including this model, we both receiving and sending of knowledge
test for the possibility of the three job char- (coefficients of 0.50 and 0.57, respectively, and
acteristics’ affecting other types of motiva- p < 0.01 for both). Introjected motivation has
tion than we hypothesized in the theoretical a significant and positive impact on sending
model (Figure 1). In addition, model 3 (in of knowledge (coefficient: 0.21, p < 0.05) but
Table IV) presents the goodness-of-fit of the is insignificant in relation to receiving of
model where the receiving and sending of knowledge. External motivation is insignifi-
knowledge are put together in one construct. cantly related to receiving knowledge but un-
As Table IV shows, only the measurement expectedly negatively related to sending it
model and the theoretical model really meet (coefficient: –0.17). The negative relation, how-
the required values for the different good- ever, is only marginally significant at p < 0.10.
ness-of-fit statistics. The model where the In sum, H1 concerning intrinsic motivation is
receiving and sending of knowledge are col- confirmed, and H2 on introjected motivation
lapsed into one measure (model 3) obtained is somewhat supported (for sending of knowl-

Autonomy 0.39*** Intrinsic 0.50***

motivation
0.57*** Receiving of
0.15*
-0.08 knowledge
-0.06
Task 0.23*** Introjected
Identity motivation
0.21**
0.27***
0.03 Sending of
knowledge
0.20*** External
Feedback -0.17*
motivation
*, **, ***: 10%, 5%, and 1% level of significance, respectively
FIGURE 2. Empirical Model
edge), while H3 on external motivation must We find that the three types of motivation
be rejected. have strong effects on both sending and re-
The R2 values of receiving and sending ceiving knowledge. In particular, all
knowledge are 0.25 and 0.32, respectively. three motivation types affect the ex-
This is a further indication that individual tent of knowledge sent to colleagues. The nuanced
motivation explains a large proportion of While intrinsic motivation has a very
the knowledge-sharing behavior. Further- strong and positive impact on knowl- understanding of
more, as we expected, some types of moti- edge sent, introjected motivation has motivation in this
vation are more closely correlated than a slightly weaker, yet positive, effect.
others. Introjected motivation is positively External motivation, on the other research reveals
(and significantly) correlated with both in- hand, is negatively related to the ex-
trinsic motivation and external motivation, tent of knowledge sent to colleagues. that the type of
while the two ends of the scale—that is, in- The reason might be that individuals
motivation fostered
trinsic and extrinsic motivation—are uncor- who engage in knowledge sharing for
related. external reasons strategize more on matters, and all
their knowledge sharing by sharing
the amount of knowledge required to types of motivation
Concluding Discussion
obtain external rewards. The extent
are not equally
The main purpose of this research is to further of knowledge the focal employee re-
our understanding of how different aspects of ceives, however, is only significantly desirable for
job design foster different types of individual influenced by intrinsic motivation
motivation (i.e., intrinsic, introjected, and toward knowledge sharing; as we ex- knowledge sharing.
external motivation), as well as how these pected, this effect is positive.
motivation types influence employees’ knowl-
edge-sharing behavior. Our results show that
Future Research
(1) job autonomy increases employees’ intrin-
sic motivation toward knowledge sharing, (2) The nuanced understanding of motivation in
task identity is positively linked to introjected this research reveals that the type of motiva-
motivation toward knowledge sharing, and tion fostered matters, and all types of motiva-
(3) feedback on the job has a positive impact tion are not equally desirable for knowledge
on employees’ external motivation to engage sharing. This has implications for future re-
in knowledge sharing. search on job design. Recent research has

applied job design theory in related areas these job characteristics. We did test, how-
such as proactive (Fuller et al., 2006) and pro- ever, for alternative specifications of the model
social behavior (Grant, 2007, 2008b). We like model 2 presented in Table IV, where each
have further expanded its domain of applica- of the three job design variables is linked to
tion to knowledge sharing. Spe- all three types of motivation. In this model,
cifically, by elaborating on job only the three proposed relationships be-
We encourage
characteristics conducive to moti- tween job characteristics and motivation (H4
management to vation toward knowledge sharing, through H6) were significant, while all the
we take steps in the direction of other (six relationships) were insignificant. In
take extra care reorienting job design research the same line, none of the other alternative
toward the emerging knowledge models we tested indicated significant joint
when designing governance approach (Foss, 2007; effects among the job characteristics.
jobs to incorporate Grandori, 1997, 2001). Further, Hackman and Oldham’s (1980)
Grant (2007) argues that fu- model includes two additional job character-
external feedback ture studies should consider going istics our study does not account for—namely,
beyond a rather narrow, limited skill variety and task significance—that could
mechanisms. set of job characteristics defined possibly have strengthened our model. We
largely by Hackman and Oldham’s suspect that these constructs would have
In general, it is
(1980) model. We concur and call yielded even more importance to intrinsic
important that for future research that provides a and introjected motivation relative to exter-
deeper understanding of a wider nal motivation, placing emphasis on the
employees do not set of job characteristics. It would motivation types we already find most im-
be beneficial if such research portant. For example, Grant (2007) links task
have a sense of
included more social or interper- significance to prosocial motivation, which
external pressure or sonal aspects of the job, for he in a later study argues may be an intro-
example, to reflect better the team- jected form of motivation (Grant, 2008b). On
being manipulated based aspects that characterize the other hand, skill variety is linked to chal-
many jobs today. While research- lenges and feelings of competence, which are
because this may
ers increasingly include constructs often associated with intrinsic motivation
have negative such as task interdependence (Gagné & Deci, 2005; Ryan & Deci, 2000).
(Guzzo & Shea, 1992; Langfred, We relied on a cross-sectional design and
effects on the 2007) and relational mechanisms perceptual measures, which provide us with
(Grant, 2007; 2008a; 2008b) in re- “fine-grained measures of variables that are
engagement in search models, more work is otherwise difficult to measure” (Haas & Han-
knowledge sharing. needed to understand the effects sen, 2007, p. 1150). It would be useful in the
such job design aspects have on future, however, to combine data from mul-
motivation to share knowledge. tiple sources to develop more elaborate mea-
Motivation types such as reciprocity (Fehr & sures. Finally, we only examined individuals
Fischbacher, 2002) and obligation-based (Lin- working for one company located in Den-
denberg, 2001) motivation are relevant here, mark. The findings we have reported here
as the more social and team-based character- thus may be a reflection of company- and
istics possibly are linked to motivation types country-specific attributes. There is a need for
like these. further empirical studies using individual
data gathered from a wider variety of firms to
generalize our findings further. Still, we are
Limitations
confident that the model developed and
We have identified the unique and isolated tested in this research provides evidence on
effects of each of the three job characteristics the role of HRM practices in general, and job
included in our model. This is a simplification design in particular, in governing individual
of the Hackman and Oldham (1980) model motivation to share knowledge. To test this
because we do not look at the joint effects of proposition further, we need to consider a

wider range of HRM practices and introduce ensure employees feel autonomous in areas
more individual-level variables in addition such as scheduling work, making decisions,
to motivation (e.g., ability and opportunity; and determining how to do the job. If man-
cf. Guest, 1997). agement wants to enhance introjected moti-
vation, on the other hand, it should make
sure employees identify with the tasks per-
Implications for Managers formed through job enrichment that ensures
Since different job characteristics have the one person is responsible for producing a
potential to enhance certain types of whole product or entire service and has a vis-
motivation, management needs to consider ible outcome.
carefully how it designs specific jobs. Job With respect to specific HRM practices,
characteristics theory implies designing jobs we encourage management to take extra
that introduce three job characteristics to the care when designing jobs to incorporate
maximum degree possible. Our findings call external feedback mechanisms. In general,
for differentiation. For jobs where the success it is important that employees do not have
of employee performance depends on receiv- a sense of external pressure or being ma-
ing knowledge from other parts of the nipulated because this may have negative
organization, management must grant the effects on the engagement in knowledge
employee sufficient autonomy so intrinsic sharing. Our data indicated that formal
motivation toward knowledge sharing is recognition, performance evaluation, and
strengthened. For jobs where knowledge rel- feedback by employees’ supervisor may have
evant to others is created, management this negative effect. Management thus
should foster either intrinsic or introjected needs to take special care when crafting
motivation to enhance the likelihood of em- such feedback practices and make sure that
ployees’ sending knowledge to colleagues. employees perceive them as informative
Further, our results show that to stimulate and development oriented rather than con-
intrinsic motivation, management should trolling.
NICOLAI J. FOSS is professor of organization and strategy at the Copenhagen Business

School and the Norwegian School of Economics and Business Administration and a
Director of the CBS Center for Strategic Management and Globalization. Nicolai’s research
interests are the theory of the firm, knowledge governance, and strategic management.
His work has been published in journals such as the Strategic Management Journal,
Organization Science, and Journal of International Business Studies.
DANA B. MINBAEVA (Ph.D., Copenhagen Business School) is an associate professor in

strategic human resource management at the Center for Strategic Management and
Globalization, Copenhagen Business School, Denmark. Her current research focuses on
strategic international HRM and knowledge sharing and transfer in multinational corpo-
rations. She has published in Journal of International Business Studies, Management
International Review, International Journal of Human Resource Management, Interna-
tional Business Review, Personnel Review, Employee Relations, and European Journal
of International Management. In addition to teaching in various graduate programs as
well as executive development programs at Copenhagen Business School, she has con-
ducted training programs and consulted for a number of large multinational corpora-
tions and international organizations in Kazakhstan, Lithuania, and Kyrgyzstan.
TORBEN PEDERSEN is professor at Copenhagen Business School’s Center for Strategic

Management and Globalization. He is Vice President of Academy of International Busi-
ness, AIB 2009 Program Chair, and Vice Chairman of the board of European International

Business Academy. Pedersen’s research interests include the interface among strategy,
organization theory, and institutional economics. He has published more than 50 articles
and books concerning managerial and strategic aspects of globalization. His research
has appeared in prominent journals such as the Journal of International Business Stud-
ies, Strategic Management Journal, Journal of Corporate Finance, and Management In-
ternational Review.
MIA REINHOLT is assistant professor of management at the Center for Strategic Manage-
ment and Globalization, Copenhagen Business School. She received her Ph.D., in strate-
gic management at Copenhagen Business School. Her current research interests include
motivation management, knowledge management, employee knowledge networks, and
the microfoundations of organizations at a general level.
References Cabrera, E. F., & Cabrera, A. (2005). Fostering knowl-

edge sharing through people management
Amabile, T. M. (1993). Motivational synergy: Toward practices. International Journal of Human Resource
new conceptualizations of intrinsic and extrinsic Management, 16, 720–735.
motivation in the workplace. Human Resource
Chatman, J., & Flynn, F. (2005). Full-cycle micro-
Management Review, 3, 185–201.
organizational behavior research. Organization
Aoki, M. (1986). Horizontal vs. vertical information Science, 16, 437–447.
structures of the firm. American Economic Review,
Clegg, C., & Spencer, C. (2007). A circular and dynamic
76, 971–983.
model of the process of job design. Journal of
Argote, L., & Ingram, P. (2000). Knowledge transfer: Occupational and Organizational Psychology, 80,
A basis for competitive advantage in firms. Organi- 321–339.
zational Behavior and Human Decision Processes,
Cohen, W. M., & Levinthal, D. A. (1990). Absorptive ca-
82, 150–169.
pacity: A new perspective on learning and innova-
Baard, P. P., Deci, E. L., & Ryan, R. M. (2004). Intrinsic tion. Administrative Science Quarterly, 35, 128–153.
need satisfaction: A motivational basis of perform- Coleman, J. S. 1990. Foundations of social theory.
ance and well-being in two work settings. Journal Cambridge, MA: Belknap Press of Harvard
of Applied Social Psychology, 34, 2045–2068. University Press.
Beugelsdijk, S. (2008). Strategic human resource prac- Davenport, T. H., & Prusak, L. (1998). Working knowl-
tices and product innovation. Organization Studies, edge: How organizations manage what they know.
29(6), 821–847. Boston: Harvard Business School Press.
Bock, G. W., Zmud, R. W., Kim, Y. G., & Lee, J. N. Deci, E. L. (1971). Effects of externally mediated re-
(2005). Behavioral intention formation in knowl- wards on intrinsic motivation. Journal of Personal-
edge sharing: Examining the role of extrinsic ity and Social Psychology, 18, 105–115.
motivators, social-psychological forces, and organi-
zational climate. MIS Quarterly, 29, 87–111. Deci, E. L., Koestner, R., & Ryan, R. M. (1999). A meta-
analytic review of experiments examining the
Bollen, K. A. (1989). Structural equations with latent effects of extrinsic rewards on intrinsic motivation.
variables. New York: Wiley. Psychological Bulletin, 125, 627–668.
Burgess, D. (2005). What motivates employees Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation
to transfer knowledge outside their work unit? and self-determination in human behavior. New
Journal of Business Communication, 42, York: Plenum Press.
324–348.
Deci, E. L., & Ryan, R. M. (2000). The “what” and
Cabrera, A., & Cabrera, E. (2002). Knowledge sharing “why” of goal pursuits: Human needs and the
dilemmas. Organization Studies, 23, 687–710. self-determination of behavior. Psychological
Cabrera, A., Collins, W. C., & Salgado, J. F. (2006). Inquiry, 11, 227–268.
Determinants of individual engagement in knowl- Eisenhardt, K., & Graebner, M. (2007). Theory building
edge sharing. International Journal of Human from cases: Opportunities and challenges. Acad-
Resource Management, 17, 245–264. emy of Management Journal, 50, 25–32.

Elster, J. (1989). Nuts and bolts for the social sciences. Grandori, A. (2001). Neither hierarchy nor identity:
Cambridge: Cambridge University Press. Knowledge governance mechanisms and the
theory of the firm. Journal of Management and
Fehr, E., & Fischbacher, U. (2002). Why social prefer-
Governance, 5, 381–399.
ences matter: The impact of non-selfish motives on
competition, cooperation and incentives. Economic Grant, A. (2007). Relational job design and the motiva-
Journal, 112, C1–C33. tion to make a prosocial difference. Academy of
Management Review, 32(2), 393–417.
Felin, T., & Foss, N. J. (2006). Individuals and organiza-
tions: Thoughts on a micro-foundations project for Grant, A. M. (2008a). The significance of task sig-
strategic management and organizational analysis. nificance: Job performance effects, relational
Research Methodology in Strategy and Manage- mechanisms, and boundary conditions. Journal of
ment, 3, 253–288. Applied Psychology, 93, 108–124.
Felin, T., & Hesterly, W. (2007). The knowledge-based Grant, A. M. (2008b). Does intrinsic motivation fuel the
view, nested heterogeneity, and new value crea- prosocial fire? Motivational synergy in predicting
tion: Philosophical considerations on the locus of persistence, performance, and productivity.
knowledge. Academy of Management Review, 32, Journal of Applied Psychology, 93, 48–58.
195–218.
Guest, D. (1997). Human resource management and
Fornell, C., & Larcker, D. (1981). Evaluating structural performance: A review and research agenda. Inter-
equation models with unobservable variables and national Journal of Human Resource Management,
measurement error. Journal of Marketing Research, 8, 263–276.
18, 39–50.
Gupta, A. K., & Govindarajan, V. (2000). Knowledge
Foss, N. J. (2007). The emerging knowledge govern- flows within multinational corporations. Strategic
ance approach: Challenges and characteristics. Management Journal, 21, 473–496.
Organization, 14, 29–52.
Guzzo, R. A., & Shea, G. P. (1992). Group performance
Frese, M., Kring, W., Soose, A., & Zempel, J. (1996). and intergroup relations in organizations. In M. D.
Personal initiative at work: Differences between Dunnette & L. M. Hough (Eds.), Handbook of
East and West Germany. Academy of Management industrial and organizational psychology (Vol. 3,
Journal, 39, 37–63. pp. 269–313). Palo Alto, CA: Psychological Press.
Frey, B. S. (1997). Not just for the money: An economic Haas, M., & Hansen, M. (2007). Different knowledge,
theory of personal motivation. Cheltenham, UK: different benefits: Toward a productivity perspec-
Edward Elgar. tive on knowledge sharing in organizations. Strate-
Fried, Y., Oldham, G., & Cummings, A. (2001). Job gic Management Journal, 28, 1133–1153.
design. In M. Warner (Ed.), The international Hackman, J. R., & Oldham, G. R. (1975). Development
encyclopedia of business management. London: of the job diagnostic survey. Journal of Applied
Thomson. Psychology, 60, 159–170.
Fuller, J. B., Marler, L. E., & Hester, K. (2006). Promot- Hackman, J. R., & Oldham, G. R. (1976). Motivation
ing felt responsibility for constructive change and through the design of work: Test of a theory.
proactive behavior: Exploring aspects of an elabo- Organizational Behavior and Human Performance,
rated model of work design. Journal of Organiza- 16, 250–279.
tional Behavior, 27, 1089–1120.
Hackman, J. R., & Oldham, G. R. (1980). Work rede-
Gagné, M. (2003). The role of autonomy support and
sign. Reading, MA: Addision-Wesley.
autonomy orientation in prosocial behavior en-
gagement. Motivation and Emotion, 27, 199–223. Hair, J., Anderson, R., Tatham, R., & Black, W. (1995).
Multivariate data analysis with readings (4th ed.).
Gagné, M., & Deci, E. L. (2005). Self-determination
Englewood Cliffs, NJ: Prentice-Hall.
theory and work motivation. Journal of Organiza-
tional Behavior, 26, 331–362. Harackiewicz, J. M., Manderlink, G., & Sansone, C.
(1984). Rewarding pinball wizardry: The effects of
Gerbing, D., & Anderson, J. (1988). An updated
evaluation on intrinsic interest. Journal of Person-
paradigm for scale development incorporating
ality and Social Psychology, 47, 287–300.
unidimensionality and its assessment. Journal of
Marketing Research, 25, 186–192. Higgins, E. T. (1997). Beyond pleasure and pain.
American Psychologist, 52, 1280–1300.
Grandori, A. (1997). Governance structures, coordina-
tion mechanisms and cognitive models. Journal of Hooff, B. van den, & de Leeuw van Weenen, F. (2004).
Management and Governance, 1, 29–42. Committed to share: Commitment and MC use as

antecedents of knowledge sharing. Knowledge and Lepak, D. P., & Snell, S. A. (1999). The human resource
Process Management, 11, 13–24. architecture: Toward a theory of human capital
allocation and development. Academy of
Howard, G. (1994). Why do people say nasty things
Management Review, 24, 31–48.
about self-reports? Journal of Organizational
Behavior, 15, 399–404. Lepak, D. P., & Snell, S. A. (2002). Examining the
human resource architecture: The relationships
Jackson, S., Chuang, C., Harden, E., & Jiang, Y. (2006).
among human capital, employment, and human
Toward developing human resource management
resource configurations. Journal of Management,
systems for knowledge-intensive teamwork.
28, 517–543.
Research in Personnel and Human Resources
Management, 25, 27–70. Lepper, M. R., Greene, D., & Nisbett, R. E. (1973).
Undermining children’s intrinsic interest with
Joreskog, K., & Sorbom, D. (1993). LISREL 8: Structur-
extrinsic reward: A test of the “overjustification”
al equation modeling with the SIMPLIS command
hypothesis. Journal of Personality and Social
language. Hillsdale, NJ: Lawrence Erlbaum.
Psychology, 28, 129–137.
Kang, S. C., Morris, S., & Snell, S. A. (2007).
Lin, H. (2007). Effects of extrinsic and intrinsic motiva-
Relational archetypes, organizational learning,
tion on employee knowledge sharing intentions.
and value creation: Extending the human resource
Journal of Information Science, 33, 135–149.
architecture. Academy of Management Review,
32, 236–256. Lindenberg, S. (2001). Intrinsic motivation in a new
light. Kyklos, 54, 317–342.
Koestner, R., & Losier, G. F. (2002). Distinguishing
three ways of being highly motivated: A closer Locke, E. A., & Latham, G. P. (2002). Building a
look at introjection, identification, and intrinsic practically useful theory of goal-setting and task
motivation. In E. L. Deci & R. M. Ryan (Eds.), motivation: A 35-year odyssey. American
Handbook of self-determination research Psychologist, 57, 705–717.
(pp. 101–121). Rochester, NY: University of Lyles, M., & Salk, J. (1996). Knowledge acquisition
Rochester Press. from foreign parents in international joint ventures:
Kogut, B., & Zander U. (1992). Knowledge of the firm, An empirical examination in the Hungarian context
combinative capabilities, and the replication of [Special issue]. Journal of International Business
technology. Organization Science, 3, 383–397. Studies, 877–903.
Kruglanski, A. W., Friedman, I., & Zeevi, G. (1971). The Minbaeva, D. B. (2005). HRM practices and MNC
effects of extrinsic incentives on some qualitative knowledge transfer. Personnel Review, 34(1):
aspects of task performance. Journal of Personal- 125–144.
ity, 39, 606–617. Minbaeva, D., Pedersen, T., Björkman, I., Fey, C., &
Langfred, C. W. (2007). The downside of self- Park, H. J. (2003). MNC knowledge transfer, sub-
management: A longitudinal study of the effects sidiary absorptive capacity, and HRM. Journal of
of conflict on trust, autonomy, and task interde- International Business Studies 34, 586–599.
pendence in self-managing teams. Academy of Nickerson, J., & Zenger, T. (2004). A knowledge-based
Management Journal, 50, 885–900. theory of the firm: The problem-solving perspec-
Latham, G. P., & Pinder, C. C. (2005). Work motivation tive. Organization Science, 15, 617–632.
theory and research at the dawn of the twenty-first O’Dell, C., & Grayson, C. J. (1998). If only we knew
century. Annual Review of Psychology, 56, what we know: Identification and transfer of inter-
485–516. nal best practices. California Management Review,
Laursen, K., & Foss, N. J. (2003). New HRM practices, 40, 154–174.
complementarities, and the impact on innovation Osterloh, M., & Frey, B. (2000). Motivation, knowledge
performance. Cambridge Journal of Economics, 27, transfer and organizational form. Organization
243–263. Science, 11, 538–550.
Lawler, E. E., III, Hackman, R. J., & Kaufman, S. (1973). Parker, S., & Wall, T. (1998). Job and work design:
Effects of job redesign: A field experiment. Journal Organizing work to promote well-being and
of Applied Social Psychology, 3, 49–62. effectiveness. London: Sage.
Lawrence, P. R., & Lorsch, J. W. (1967). Organization Parker, S. K., Wall, T. D., & Jackson, P. R. (1997).
and environment: Managing differentiation and That’s not my job: Developing flexible employee
integration. Boston: Harvard Business School orientations. Academy of Management Journal,
Press. 40, 899–929.

Podsakoff, P. M., MacKenzie, S. B., & Podsakoff, N. P. Simonin, B. L. (1999). Transfer of marketing
(2003). Common method bias in behavioral know-how in international strategic alliances.
research: A critical review of the literature and Journal of International Business Studies, 30,
recommended remedies. Journal of Applied 463–490.
Psychology, 88, 879–903.
Sims, H., Szilagyi, A., & Keller, R. (1976). The measure-
Podsakoff, P. M., & Organ, D. (1986). Self-reports in ment of job characteristics. Academy of Manage-
organizational research: Problems and prospects. ment Journal, 19, 195–212.
Journal of Management, 12, 531–544.
Spector, P. (1994). Using self-report questionnaires in
Quigley, N. R., Tesluk, P. E., Locke, E. A., & Bartol, K. M. OB research: A comment on the use of a controver-
(2007). A multilevel investigation of the motivation- sial method. Journal of Organizational Behavior,
al mechanisms underlying knowledge sharing and 15, 385–392.
performance. Organization Science, 18, 71–88.
Szulanski, G. (1996). Exploiting internal stickiness:
Ryan, R. M. (1995). Psychological needs and the Impediments to the transfer of best practice.
facilitation of integrative processes. Journal of Strategic Management Journal, 17, 27–43.
Personality, 63, 397–427.
Tsai, W., & Ghoshal, S. (1998). Social capital and value
Ryan, R. M., & Connell, J. P. (1989). Perceived locus of creation: The role of intrafirm networks. Academy
causality and internalization: Examining reasons of Management Journal, 41, 464–476.
for acting in two domains. Journal of Personality
Vallerand, R. J., & Bissonnette, R. (1992). Intrinsic,
and Social Psychology, 57, 749–761.
extrinsic, and amotivational styles as predictors of
Ryan, R. M., & Deci, E. L. (2000). Self-determination behavior: A prospective study. Journal of Personal-
theory and the facilitation of intrinsic motivation, ity, 60, 599–620.
social development, and well-being. American
Vansteenkiste, M., Simons, J., Lens, W., Sheldon,
Psychologist, 55, 68–78.
K. M., & Deci, E. L. (2004). Motivating learning,
Ryan, R. M., Koestner, R., & Deci, E. L. (1991). Ego- performance, and persistence: The synergistic
involved persistence: When free-choice behavior effects of intrinsic goal contents and autonomy-
is not intrinsically motivated. Motivation and supportive contexts. Journal of Personality and
Emotion, 15, 185–205. Social Psychology, 87, 246–260.
Salancik, G., & Pfeffer, J. (1977). An examination of Vroom, V. H. (1964). Work and motivation. New York:
need-satisfaction models of job attitudes. Adminis- Wiley.
trative Science Quarterly, 22, 427–456.
Wright & Niishi, 2006. Strategic HRM and Organi-
Siggelkow, N. (2007). Persuasion with case zational behaviour: Integrating multiple levels of
studies. Academy of Management Journal, 50, analysis. Working Paper 06-05. CAHRS Working
20–24. Paper Series, Cornell University, NY.

2009 Fossetal HRM PDF

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

2009 Fossetal HRM PDF

Transféré par

Droits d'auteur :

Formats disponibles

ENCOURAGING KNOWLEDGE

SHARING AMONG EMPLOYEES:

Keywords: job design, job characteristics, types of motivation, knowledge

Introduction therefore be an important design variable for

J frequently researched Human Re-

Human Resource Management DOI: 10.1002/hrm

Human Resource Management DOI: 10.1002/hrm

Human Resource Management DOI: 10.1002/hrm

FIGURE 1. Theoretical Model

Human Resource Management DOI: 10.1002/hrm

Human Resource Management DOI: 10.1002/hrm

Human Resource Management DOI: 10.1002/hrm

Methods We collected the data using an administered

Human Resource Management DOI: 10.1002/hrm

Human Resource Management DOI: 10.1002/hrm

The Motivation to Share Knowledge jected, and external motivation. We adopted

TABLE I Response Distribution

Human Resource Management DOI: 10.1002/hrm

TABLE II Constructs and Items*

Human Resource Management DOI: 10.1002/hrm

Human Resource Management DOI: 10.1002/hrm

Human Resource Management DOI: 10.1002/hrm

TABLE III Continued.

Human Resource Management DOI: 10.1002/hrm

TABLE IV Goodness-of-Fit Statistics for Three Competing Specifications of the Model

Human Resource Management DOI: 10.1002/hrm

Human Resource Management DOI: 10.1002/hrm

Autonomy 0.39*** Intrinsic 0.50***

*, **, ***: 10%, 5%, and 1% level of significance, respectively

FIGURE 2. Empirical Model

Human Resource Management DOI: 10.1002/hrm

Human Resource Management DOI: 10.1002/hrm

NICOLAI J. FOSS is professor of organization and strategy at the Copenhagen Business

DANA B. MINBAEVA (Ph.D., Copenhagen Business School) is an associate professor in

TORBEN PEDERSEN is professor at Copenhagen Business School’s Center for Strategic

Human Resource Management DOI: 10.1002/hrm

References Cabrera, E. F., & Cabrera, A. (2005). Fostering knowl-

Human Resource Management DOI: 10.1002/hrm

Human Resource Management DOI: 10.1002/hrm

Human Resource Management DOI: 10.1002/hrm

Human Resource Management DOI: 10.1002/hrm

Vous aimerez peut-être aussi

Autonomy 0.39* Intrinsic 0.50*

, , : 10%, 5%, and 1% level of significance, respectively