Vous êtes sur la page 1sur 11

Infant Behavior & Development 36 (2013) 199209

Contents lists available at SciVerse ScienceDirect

Infant Behavior and Development

Gender and discipline in 512-month-old infants: A longitudinal


study
Richard Evan Ahl a,1 , Anne Fausto-Sterling b, , Cynthia Garca-Coll c , Ronald Seifer d
a
b
c
d

Brown University, United States


Department of Molecular Biology, Cell Biology and Biochemistry, Brown University, United States
Departments of Education and Cognitive, Linguistic Sciences and Psychology, Brown University, United States
Department of Psychiatry and Human Behavior, Brown University, United States

a r t i c l e

i n f o

Article history:
Received 31 August 2012
Received in revised form
28 November 2012
Accepted 12 January 2013
Keywords:
Gender
Discipline
Motherinfant interaction
Infancy

a b s t r a c t
We examined the effects of infant age and gender on the behaviors of infants and mothers
during discipline interactions using longitudinal, naturalistic, home-based, taped observations of 16 motherinfant dyads (eight males and eight females). These observations were
conducted between the child ages of 5 and 12 months and used a devised Maternal Discipline Coding System to code for the occurrence of discipline events. During discipline
interactions, mothers vocalized longer, used harsher tones, and used more explanations
with older compared to younger infants. Male infants were more likely than female infants
to cry or whine during discipline events. Mothers of male infants used longer vocalizations,
more words, and more affectionate terms than mothers of female infants. Male infants
were more difcult during discipline interactions than female infants, but it appeared that
mothers of males responded to this difculty by using milder discipline techniques.
2013 Elsevier Inc. All rights reserved.

1. Gender and discipline in infancy


In their groundbreaking 1974 book on sex differences, Maccoby and Jacklin cited sex differences in aggression as one of
the few robust ndings in the psychological literature. Since this important work much attention has focused on different
forms of aggression between girls and boys (Ostrov & Keating, 2004). Generally, it is believed that these different types of
aggression, which have the potential to become disruptive and even anti-social in extreme manifestations, may emerge from
a variety of parenting characteristics, including ones that also promote insecure attachment behaviors. Despite a considerable
literature on sex differences in aggression and its relationship to attachment and on the potential negative consequences of
overly internalizing and overly externalizing behavior, there are very few studies of parent child disciplinary interactions
(a) with infants and (b) in which possible gender differences are examined (Casas et al., 2006; Kawabata, Alink, Tseng, van
Ijzendoorn, & Crick, 2011; Nelson, Hart, Yang, Olsen, & Jin, 2006). Here we report the results of a small study of mother infant
disciplinary interactions observed regularly over a several month period.
The literature on gender and discipline is far from clear. Straus et al. (1998) found that mothers generally discipline
young children more harshly than fathers, primarily due to the greater time mothers spend caring for their children. In a
survey-based study on parents of children between 4 and 35 months of age, Regalado, Sareen, Inkelas, Wissow, and Halfon

This work was supported by funding from the Ford Foundation and a Brown University summer fellowship for Richard Ahl.
Corresponding author at: Department of Molecular Biology, Cell Biology and Biochemistry, Box G, Brown University, Providence, RI 02912,
United States. Tel.: +1 401 863 2109.
E-mail addresses: Anne Fausto-Sterling@brown.edu, afs@brown.edu (A. Fausto-Sterling).
1
Current address: Framingham Heart Study, 73 Mt. Wayte Avenue, Suite 2, Framingham, MA 01702-5827, United States.
0163-6383/$ see front matter 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.infbeh.2013.01.005

200

R.E. Ahl et al. / Infant Behavior & Development 36 (2013) 199209

(2004) found that the sex of the child had no effect on parentally reported types of discipline. Socolar, Savage, Keyes-Elstein,
and Evans (2005) also found that child gender did not inuence parental discipline in their study on parental discipline of
1219-month-old children.
In an observation-based study of toddlers (14 months) and their mothers, Power and Chapieski (1986) found that the sex
of the child had no signicant effect on the childs rate of misbehavior (playing with forbidden objects) or the mothers use of
discipline. Their review of similar studies found few sex differences in measures of toddler compliance and misbehavior. In
an observation-based study, Kochanska, Kuczynski, and Radke-Yarrow (1989) found that the sex of the child (1644 months)
had no bearing on maternal use of child management methods. More recent studies by Kochanska, however, have found
sex differences (Kochanska, 2001; Kochanska, Aksan, & Nichols, 2003; Kochanska, Coy, & Murray, 2001; Kochanska, Coy,
Tjebkes, & Husarek, 1998; Kochanska, Tjebkes, & Forman, 1998).
Using both survey and observational measures of maternal discipline and child misbehavior in 13-year-olds, van Zeijl
et al. (2007) found that mothers of boys and mothers of girls did not differ in their use of discipline strategies. However,
boys showed higher rates of physical aggression than girls. Lytton and Romney (1991) conducted a meta-analysis of 172
studies on child gender and parent behavior in an attempt to determine whether parents make systematic differences in
their rearing of boys and girls (p. 267). The study found few signicant differences, but there was a slight tendency for
parents to use more physical punishment on boys than girls.
Miner and Clarke-Stewarts (2008) ndings of sex differences in child externalizing behavior raise important methodological issues. Mothers who did not use harsh discipline and had higher maternal sensitivity ratings were effective at
reducing levels of child externalizing behavior at later ages, an effect that was especially strong for boys. However, the nding of sex differences in rates of child externalizing behavior depended upon the identity of the reporter, as teacher ratings
were signicantly higher for the boys than for the girls. There was no signicant difference between the ratings of mothers
of boys and the ratings of mothers of girls. The use of direct observations, rather than adult reporting methods, could help
determine whether the source of such differences stems from informant bias or contextual differences in child behavior.
The research of McFadyen-Ketchum, Bates, Dodge, and Pettit (1996) suggests that maternal discipline effects may have a
different trajectory for boys and girls. The study investigated whether levels of maternal coercion and non-affection during
discipline interactions predicted levels of child disruption and aggression from kindergarten to third grade. Although a
maternal behavior prole of high coercion and low affection was linked to increased levels of aggression during kindergarten
for boys and girls, the effects of maternal behavior on changes in the childs behavioral prole depended on the childs sex.
For boys, a high coercion/low affection behavior prole was associated with increased aggression over time. For girls, higher
levels of coercion actually predicted decreased aggression over time, and there was no relationship between maternal
affection and aggression changes.
Kochanska et al. (1989, 2003) observed boys and girls in laboratory settings with their mothers during the ages of 14, 22, 33
and 45 months. The observation scenarios consisted of do (toy cleanup) and dont (refrain from touching attractive toys)
tasks. They found that mothers of boys used more power assertion than mothers of girls. Girls demonstrated higher rates
of empathy and maternal imitation than boys. Another study focused on rates of child compliance in response to maternal
requests and prohibitions (Kochanska et al., 2001). For the do task, girls showed higher rates of committed compliance
at 14 months, but rates for both sexes were similar for the subsequent age groups. For the dont task, the sex differences
were more pronounced: girls showed signicantly higher rates of committed compliance at every age group.
To sum up, some studies (Kochanska et al., 1989; Power & Chapieski, 1986; Regalado et al., 2004; Socolar et al., 2005; van
Zeijl et al., 2007) have found that mothers of boys and girls have similar patterns in overall discipline practices, while others
(Kochanska et al., 2003; Laumann, Michael, & Gagnon, 1994) have found differences. The direction of causality in studies
that nd sex differences is usually unclear, i.e. whether mothers use more harsh discipline on boys because they misbehave
more frequently or whether the harsh discipline causes more misbehavior, or both.

1.1. Using dynamic systems theory to study gender and discipline


Many studies on maternal discipline and child behavior only use survey data, making it difcult to determine what an
actual discipline event looks like, the complexities of the dyadic exchange, the contextual factors that inuence how the
event unfolds, etc. Moreover, much of the observational research has been conducted using experimentally manipulated
scenarios designed to elicit discipline events (i.e. the do/dont tasks). While these approaches are important, they do not
necessarily clarify the types of discipline that occur in everyday situations. Exceptions include the work of Hollenstein and
colleagues who use state space grids to analyze moment by moment dynamics of disciplinary interactions between mothers
and children (Granic, Hollenstein, Dishion, & Patterson, 2003; Hollenstein, Granic, Stoolmiller, & Snyder, 2004).
Many studies restrict observations to a small number of sessions conducted several months apart. While less labor and
time-intensive, this approach may miss subtle changes in discipline interactions over time. Finally, it is important to conduct
research on the discipline of very young children. Results from survey-based studies have shown that parents do discipline
their children between the ages of 5 and 12 months, but very little observation-based work has examined this age group.
It is unclear, for instance, how discipline practices at 5 months differ from discipline practices at 12 months. The greatest
potential to set healthy patterns of discipline could be during early infancy (see Ravn et al., 2011, 2012) before the onset of
discipline challenges that are created during the childs second year of life. For these methodological and theoretical reasons

R.E. Ahl et al. / Infant Behavior & Development 36 (2013) 199209

201

we examined the effects of infant age and gender on the behaviors of infants and mothers during discipline interactions
using longitudinal, naturalistic, home-based, taped observations of 16 motherinfant dyads.
2. Methods
2.1. Participants
The participants consisted of 16 motherinfant pairs (8 male infants, 8 female infants) initially recruited for a study on
infant temperament. Families were rst contacted during the lying-in period at the university obstetrics hospital, which
accounted for more than 90%of the births in the state of Rhode Island. Direct contact was made with white, rst time
mothers who were 20 years of age or older. Subsequent contact, explanations of the project and obtaining of agreement
to participate resulted in a sample of families that was relatively afuent, intact, homogeneous, and motivated to continue
with an intensive longitudinal study commitment. Thus all of the enrolled mothers, taped in the early 1990s, were married,
white, rst-time mothers with a mean age of 29.1 (SD = 4.2). The families had an average maternal Hollingshead ISP of 48.38
(SD = 8.91), paternal Hollingshead ISP of 48.75 (SD = 15.07), and 4-factor SES of 1.81 (SD = .75) (Seifer et al., 1994). After the
coding was completed, independent samples t tests comparing the families of male infants and female infants revealed that
mothers of male infants (M = 7.13, SD = 1.25) had a signicantly higher mean occupation status (Hollingshead) than mothers
of female infants (M = 5.63, SD = 1.41), t (14) = 2.26, p = .04. Male and female infants families did not differ for any of the other
parental variables, including maternal education. Independent samples t tests revealed no signicant differences between
males and females on any of the temperament variables.
2.2. Materials
The original videotapes contained sessions that typically lasted for 1 h. In each session, the mother was asked to spend
at least 10 min in each of the following situations: (a) play with infant, (b) caretaking with infant and (c) infant alone (with
the mother supervising the infants activities but not partaking in them). Taping was conducted weekly, with some missed
weeks for each family due to vacations, illness, etc. The vast majority of taping sessions were conducted in the living rooms
or play rooms of the participants homes. A small number of taping sessions were conducted outdoors. The percentage of
outdoor observations was similar across all participants.
We analyzed 251 7-min taped observations. From each taping session, a continuous, 7-min segment of play with infant
footage was selected. Brief interruptions of the observation (i.e. answering the phone) were allowed, but if the total length
of interruptions in the clip exceeded 1 min in length, the observation was discarded. An observation would also be discarded
if other children or adults were involved in the observation, if the audio or video was damaged, or if the cohesiveness of
the observation was fundamentally altered in any other way (i.e. a switch to a caretaking task during the middle of the
observation).
The Maternal Discipline Coding System summarized in Table 1, designed by Richard Ahl, was used to analyze the footage.
The Observer XT (Noldus) was used to code the footage and perform basic analyses. SPSS was used for most of the statistical
analyses. If the assumptions of Mauchlys Test of Sphericity, Boxs Test of Equality of Covariance Matrices, and Levenes Test
Table 1
Discipline event code denitions.
Maternal codes

Denition

Affectionate terms
Description of infant misbehavior
Distraction/alternative
Explanation/reasoning
Latency
Prohibition, uses no
Prohibition, uses another
Prohibition-indirect
Understanding
Vocalization length
Vocalization tone
Word count

Mothers vocalization contains an affectionate term for her infant


Mother fully or partly describes the infants misbehavior
Mother distracts infant (toy, hugging, etc.) during or <10 s after the discipline event
Mother explains why the misbehavior is harmful, undesirable, etc.
Time lag between Infants misbehavior and mothers response
Mother says the word no during the vocalization
Mother says dont, cant, or another similar word
Mother does not use a clear prohibition during the vocalization
Mothers vocalization expresses sympathy for, or understand of, the infants thoughts/emotions
Length of mothers disciplinary vocalization, in seconds
Affective qualities of mothers vocalization prosody from 1 (very harsh) to 5 (very positive)
Number of words in mothers disciplinary vocalization

Infant codes

Denition

Danger level
Infant compliance
Infant misbehavior type

Infants misbehavior is reasonable likely to cause physical harm to the infant


Infant stops misbehaving after mother vocalizes, before she intervenes physically
Categorizes the type of infant misbehavior. The codes are: object misuse-in mouth, object
misuse-other, forbidden object, harm to mother, location, movement
Rating of infant affect during or <5 s after the disc event, from 0 (pos/neutral) to 2 (very upset)
Infant misbehavior seems imminent, but has not yet started the misbehavior
Infant is misbehaving or has nished misbehavior when the mother intervenes

Infant negative affect


Level of completion not started
Level of completion active/completed

202

R.E. Ahl et al. / Infant Behavior & Development 36 (2013) 199209

of Equality of Error Variances have not been met we note this in the text. The term mixed between-within subjects ANOVA
is synonymous with mixed-design, repeated measures ANOVA (Pallant, 2007).
2.3. Observation selection
Monthly observations were selected from the tapes collected during the infants months 512. A total of 251 7-min
observations were analyzed. The tapes were divided into 4 age groups: 56, 78, 910, and 1112 months. For 13 of the
16 dyads, four 7 min observations that met the above-described criteria, from each age group were analyzed, for a total of
16 observations per dyad. For most dyads, this meant using two observations taped during each age month. Observations
were chosen from as close to the middle of the month as possible. For some of the age groups, three dyads only had three
satisfactory observations for the given age group. Averages from the other same-gender dyads for the given age group were
calculated and then inserted in place of the missing observations.
2.4. Coding
We rst watched each 7 min segment once, noting any possible instances of child misbehavior or reactive maternal
discipline. The information gathered in this initial step was used to help create the Maternal Discipline Coding System,
which we applied in the second and third steps of the coding process. These initial observations conrmed that each dyad
had a similar distribution of observations with restricted/non-restricted infant motion, a similar distribution of locations,
and a similar distribution of free access observations.
During a second viewing, each observation was coded for the presence of discipline events using The Observer, according to the guidelines of the Maternal Discipline Coding System. In devising the discipline event code we follow Belsky,
Woodworth, and Crnic (1996), who dened Child Behavior Management Events as explicit efforts by either parent to control or direct the childs behavior (p. 562), with the exception of rule-based play (i.e. you should put the square on top of
the circle, not on top of the triangle, or no, dont put the toy truck on top of the garage; it belongs inside the garage would
not be coded as discipline events). The criteria for the discipline event code are given in Table 1. During a third viewing,
descriptive codes were added to the discipline events (i.e. did the infant display negative affect during or immediately after
the discipline event?). For simplicity and consistency, the term misbehavior will be used to describe the infants actions
that provoke disciplinary responses from their mothers.
2.5. Coding: reliability
We developed our coding scheme for infant discipline in three stages, which are described in detail in Appendix 1. As
described, during phase 2 of the development process we conducted reliability trials using a trained reliability partner who
independently double-coded 10 observations containing discipline events. These observations were of dyads used to develop
the coding scheme, but these specic observations were not used for purposes other than the reliability trials, i.e. they were
not included in the data analysis. We took this approach because we were concerned with developing a coding scheme that
was unambiguous. Reliability analysis for the codes used in this study produced kappa values ranging from .67 to .89, with
a mean kappa value of .80.
3. Results
3.1. Discipline event length
A mixed between-within subjects (repeated measures) analysis of variance was conducted to assess the effects of child
gender and age on discipline event length (in seconds). This analysis examined whether the average length of a discipline
event was longer for male or female infants and their mothers, and for infants at different ages. We hypothesized that male
infants would have longer discipline events than female infants, and that discipline event length would increase with age.
There was no signicant main effect for child age (Wilks Lambda = .87, F (3, 12) = .58, p = .64, 2p = .13) or for child gender (F
(1,14) = 2.28, p = .15, 2p = .14). However, the interaction between child gender and age was signicant (Wilks Lambda = .36, F
(3, 12) = 7.23, p = .005, 2p = .64). As depicted in Table 3, the inuence of child gender on discipline event length was moderated
by child age. While male infants and female infants had similar discipline event lengths at the ages of 56 months and 1112
months, male infants had longer discipline events than female infants at the ages of 78 months and 910 months. To clarify
the nature of the interaction, follow-up protected independent samples t tests were conducted to compare the discipline
event lengths of male infants to those of female infants at the ages of 78 months and 910 months.
For the 78 month age group, there was no signicant difference in discipline event length between male infants
(M = 12.76, SD = 11.58) and female infants (M = 8.64, SD = 2.46), t (14) = .99, p = .34. For the 910 month age group, however, male infants and their mothers (M = 10.65, SD = 2.49) had signicantly longer discipline events than female infants and
their mothers (M = 6.07, SD = 1.73), t (14) = 4.28, p = .001.

R.E. Ahl et al. / Infant Behavior & Development 36 (2013) 199209

203

Table 2
Maternal vocalization word count, by child gender and age.
Infant age

56 months

78 months

910 months

1112 months

Mothers of males: mean word count (SD)


Mothers of females: mean word count (SD)

14.63 (2.43)
10.45 (3.23)

17.38 (6.41)
12.54 (4.06)

17.76 (3.65)
8.92 (3.02)

13.60 (6.16)
13.15 (4.58)

3.2. Total discipline event duration (per observation)


A mixed between-within subjects (repeated measures) analysis of variance was conducted to assess the effects of child
gender and age on total discipline event duration (the sum of all discipline event lengths for a single observation, in seconds).
We hypothesized that male infants would have longer durations than female infants, and that older infants would have longer
durations than younger infants. Because the assumption of equality of covariance matrices was violated according to Boxs
Test of Equality of Covariance Matrices, adjustments were made. Results from months 56 and 78 were averaged, creating
a new age group of 58. Results from months 910 and 1112 were averaged, creating a new age group of 912. After this
adjustment, Boxs Test was no longer violated. However, since the assumption of Levenes Test of Equality of Error Variances
was violated, the signicance threshold was changed from .05 to .025.
The interaction between child age and gender was not signicant (F (1,14) = .50, p = .49, 2p = .04), and there was no
signicant main effect for age (F (1,14) = .25, p = .63, 2p = .02). The main effect for child gender was marginally signicant
(F (1,14) = 5.33, p = .037, 2p = .28 where the signicance threshold was dropped to p = .025) as shown in Table 4. Averaged
across all age groups, the mean discipline event total duration for male infants and their mothers was 11.56 (SD = 6.73),
nearly twice the mean of 5.82 for female infants and their mothers (SD = 2.04).
3.3. Maternal behavior
3.3.1. Maternal vocalization length during discipline events
A mixed between-within subjects (repeated measures) analysis of variance was conducted to assess the effects of child
gender and age on maternal vocalization length (in seconds). We hypothesized that mothers of females would use longer
vocalizations than mothers of males, and that mothers would use longer vocalizations as their infants grew older. The
interaction between child gender and age approached signicance (Wilks Lambda = .54, F (3, 12) = 3.40, p = .054, 2p = .46).
The effect of gender was moderated by age, being most pronounced at the 78 and 910-month age groups, as depicted in
Table 5. There was a signicant main effect for age (Wilks Lambda = .45, F (3, 12) = 4.84, p = .02, 2p = .55).
There was a signicant main effect for gender, F (1,14) = 5.01, p = .04, 2p = .26, with mothers of males having longer vocalizations than mothers of females. Because of the near signicant interaction between gender and age, follow-up protected
independent samples t tests were conducted to compare the effects of child gender on maternal vocalization length in the
78 and 910-month age groups. For the 910-month age group, the mean vocalization length for mothers of male infants
(M = 7.86, SD = 2.11) was signicantly longer than the mean vocalization length for mothers of female infants (M = 4.29,
SD = 1.53), t (14) = 3.88, p = .002. Results for the 78-month group were not signicant (t (14) = 1.67, p = .12).
3.3.2. Maternal vocalization word count
A mixed between-within subjects (repeated measures) analysis of variance was conducted to assess the effects of child
gender and age on maternal vocalization word counts (the number of words mothers used during discipline events). We
hypothesized that mothers of females would use more words than mothers of males, and that mothers would use more
words as their infants grew older.
The interaction between child gender and age was signicant (F (3, 42) = 3.11, p = .036, 2p = .18). As depicted in Table 5,
the effects of child gender were stronger during the 56, 78, and 910-month age groups than during the 1112-month
age group. The main effect for age was not signicant (F (3, 42) = 1.08, p = .37), but there was a signicant main effect for
gender (F (1, 14) = 10.42, p = .006, 2p = .427). Mothers of males used signicantly more words during their discipline-related
vocalizations and had higher word counts than mothers of females.
To clarify the nature of the interaction between child gender and age, follow-up protected independent samples t tests
were conducted to compare the effects of child gender on maternal vocalization word counts in the 56, 78 and 910month age groups. For the 78 month age group, there was no signicant word count difference between mothers of males
and mothers of females (t (14) = 1.80, p = .93). There were signicant differences in the 56 month group (t (14) = 2.92,
p = .011), and the 910 month group (t (14) = 5.27, p = .000), with mothers of males using signicantly more words during
discipline-related vocalizations than mothers of females. Means are displayed in Table 2.
3.3.3. Maternal vocalization tone
A mixed between-within subjects (repeated measures) analysis of variance was conducted to assess the effects of child
gender and age on maternal vocalization tone (the tone/affective properties of the mothers vocalization during the discipline
interaction on a scale from 1 to 5, with lower scores corresponding to harsher vocalizations). We hypothesized that mothers
of males would use harsher tones than mothers of females, and that mothers would use harsher tones as their infants grew

204

R.E. Ahl et al. / Infant Behavior & Development 36 (2013) 199209

Table 3
Effects of child gender and age on discipline event length.
Age

56 months

78 months

910 months

1112 months

Discipline event length: boys (SD)


Discipline event length: girls (SD)

7.96 (1.93)
9.18 (2.39)

12.76 (11.58)
8.64 (2.45)

10.65 (2.49)
6.07 (1.73)

9.08 (3.63)
8.93 (3.55)

Table 4
Effects of child gender and age on total discipline event duration.
Infant age

58 months

912 months

Total discipline event duration boys (SD)


Total discipline event duration girls (SD)

10.37 (7.25)
6.02 (3.11)

12.74 (9.66)
5.61 (2.13)

older. The interaction between child gender and age was not signicant (F (3, 42) = .77, p = .52, 2p = .05). The main effect for
child gender was not signicant (F (1,14) = .63, p = .44, 2p = .04).
However, there was a signicant main effect for age (F (3, 42) = 4.45, p = .008, 2p = .24). The mean tone of vocalizations
to younger infants was more positive (higher score) than the mean tone of vocalizations to older infants. Two follow-up
protected paired samples t tests were conducted to compare the mean tone used during 56 months to the mean tone
used during 1112 months, and to compare the mean tone used during 78 months to the mean tone used during 1112
months. Mean tone scores were signicantly higher (more positive) during 56 months (M = 3.36, SD = .49) than 1112
months (M = 2.84, SD = .55) (t (15) = 2.98, p = .009), and signicantly higher during 78 months (M = 3.15, SD = .39) than 1112
months (t (15) = 2.16, p = .047).
3.3.4. Use of explanation/reasoning
A mixed between-within subjects (repeated measures) analysis of variance was conducted to assess the effects of child
gender and age on the percentage of maternal vocalizations containing explanation/reasoning during discipline events. We
hypothesized that mothers of females would use more explanation/reasoning than mothers of males, and that mothers
would use more explanation/reasoning as their infants grew older. Neither the interaction between child gender and age (F
(3, 42) = 1.66, p = .19, 2p = .11) nor the main effect for child gender were signicant (F (1, 14) = .50, p = .49, 2p = .03).
However, the main effect for child age was signicant (F (3, 42) = 3.53, p = .023, 2p = .20). Three follow-up protected paired
samples t tests compared the explanation usage in different infant age groups. Mothers used signicantly more explanations
during the infant ages of 78 months (M = 33.35, SD = 26.73) than 56 months (M = 21.02, SD = 18.30), t (15) = 2.23, p = .042.
Mothers used signicantly more explanations during the infant ages of 1112 months (M = 42.24, SD = 31.38) than 56
months, t (15) = 2.75, p = .015, and more explanations during 1112 than 910 months (M = 24.66, SD = 21.48), t (15) = 2.16,
p = .048.
3.3.5. Use of affectionate terms
A mixed between-within subjects (repeated measures) analysis of variance was conducted to assess the effects of child
gender and age on the percentage of maternal vocalizations containing affectionate terms during discipline events. We
hypothesized that mothers of females would use more affectionate terms than mothers of males, and that mothers would
use fewer affectionate terms as their infants grew older.
Because the assumption of equality of covariance matrices was violated according to Boxs Test of Equality of Covariance
Matrices using four age groups, adjustments were made. Results from months 56 and 78 were averaged, creating a new age
group of 58. Results from months 910 and 1112 were averaged, creating a new age group of 912. The mixed betweenwithin subjects analysis of variance was conducted again; this time, only two age groups (58 and 912) were used instead of
four age groups. After this adjustment, the assumption of Boxs Test of Equality of Covariance Matrices was no longer violated.
However, since the assumption of Levenes Test of Equality of Error Variances was violated, the signicance threshold was
adjusted from .05 to .025.
Neither the interactions between child gender and age (F (1,14) = .001, p = .98, 2p = .00) nor the main effect for age were
signicant (F (1,14) = 3.31, p = .09, 2p = .19). The main effect for gender approached signicance (F (1,14) = 4.37, p = .055,
2p = .24). Due to the marginally signicant main effect for gender, the lack of a signicant interaction and the lack of a
signicant main effect for age, and due to the skewed distribution of the data, a MannWhitney U test was conducted to
Table 5
Effects of child gender and age on maternal vocalization length.
Infant age

56 months

78 months

910 Months

1112 Months

Mothers of males vocalization length in seconds (SD)


Mothers of females vocalization length in seconds (SD)

5.64 (1.78)
4.22 (1.44)

11.32 (10.15)
5.20 (2.07)

7.86 (2.11)
4.29 (1.53)

5.83 (2.45)
6.53 (2.52)

R.E. Ahl et al. / Infant Behavior & Development 36 (2013) 199209

205

Table 6
Effects of child age and gender on child displays of negative affect.
Infant age

58 months
912 months

Percentage of events containing infant negative affect


Males (SD)

Females (SD)

18.68 (16.06)
14.98 (11.88)

11.56 (9.06)
01.93 (3.39)

compare mothers of male infants use of affectionate terms to mothers of female infants use of affectionate terms. The use
of affectionate terms for all four age groups was averaged, and then a MannWhitney U test was performed. Mothers of
male infants (MD = 17.97, n = 8) used signicantly more affectionate terms than mothers of female infants (MD = 5.94, n = 8),
U = 13.00, z = 2.00, p = .045, r = .50.
3.4. Infant behavior
3.4.1. Display of infant negative affect
A mixed between-within subjects (repeated measures) analysis of variance was conducted to assess the effects of child
gender and age on the percentage of discipline events containing displays of infant negative affect. We hypothesized that
male infants would display negative affect at higher rates than female infants, and that infants would display negative affect
at higher rates as they grew older.
Because the assumption of equality of covariance matrices was violated according to Boxs Test of Equality of Covariance
Matrices using four age groups, adjustments were made. Results from months 56 and 78 were averaged, creating a new age
group of 58. Results from months 910 and 1112 were averaged, creating a new age group of 912. The mixed betweenwithin subjects analysis of variance was conducted again; this time, only two age groups (58 and 912) were used instead of
four age groups. After this adjustment, the assumption of Boxs Test of Equality of Covariance Matrices was no longer violated.
However, since the assumption of Levenes Test of Equality of Error Variances was violated, the signicance threshold was
adjusted from .05 to .025.
Neither the interaction between child gender and age (F (1,14) = .56, p = .47, 2p = .04) nor the main effect for age were
signicant (F (1,14) = 2.83, p = .12, 2p = .17). However, the main effect for gender was signicant (F (1,14) = 6.70, p = .021,
2p = .32), as shown in Table 6. A signicantly higher percentage of discipline events involving male infants contained displays
of negative affect than events involving female infants. Male infants were signicantly more likely to display negative affect
during/after discipline events than female infants. A MannWhitney U test was conducted to compare male infants display
of negative affect to female infants display of negative affect. The negative affect percentage for all four age groups was
averaged, and then a MannWhitney U test was performed. Male infants (MD = 13.23, n = 8) displayed signicantly more
negative affect than female infants (MD = 7.41, n = 8), U = 13.00, z = 1.99, p = .046, r = .49.
3.4.2. High and low discipline event groups
Based on their rates of discipline events during the ages of 912 months, infants were divided into two groups. The High
discipline/misbehavior group consisted of 7 infants (4 males, 3 females) who had an average of at least 1.1 discipline events
per observation during the ages of 912 months. The Low discipline/misbehavior group consisted of 7 infants (3 males,
4 females) who had an average of less than .65 discipline events per observation. (Two infants, one with an average of .88
discipline events per observation and the other with an average of 1.0 discipline events per observation, were excluded from
the grouping.)
Independent samples t tests were conducted to compare the High and Low groups on the parental SES and infant temperament variables, both of which were previously collected by Seifer et al. (1994). Fathers occupation levels were signicantly
lower for the High group than the Low group, t (12) = 2.97, p = .012. Fathers of high-discipline infants (M = 4.57, SD = 2.07)
had signicantly lower occupation statuses than fathers of low-discipline infants (M = 7.71, SD = 1.89). Infant mood scores
at both 6 (t (12) = 3.90, p = .002) and 9 months (t (12) = 2.63, p = .02) were signicantly higher for the High group than the
Low group. High mood scores indicate higher levels of sadness and fussiness. Thus, high-discipline infants (6 months:
M = 10.56, SD = .51; 9 mo: M = 10.19, SD = .56) had signicantly higher levels of sadness and fussiness at 6 and 9 months
than low-discipline infants (6 months: M = 9.56, SD = .44; 9 months: M = 9.34, SD = .65). Total infant difculty (9 months)
scores were signicantly higher for the High group than the Low group, t (12) = 2.24, p = .045. High difculty scores indicate
higher levels of temperamental difculty. High-discipline infants (M = 37.18, SD = 1.40) had signicantly higher levels of
temperamental difculty at 9 months than low-discipline infants (M = 34.77, SD = 2.48).
3.5. Discipline event latency
A mixed between-within subjects (repeated measures) analysis of variance was conducted to assess the effects of child
gender and age on discipline event latency (the amount of time that passes between the childs misbehavior and the mothers

206

R.E. Ahl et al. / Infant Behavior & Development 36 (2013) 199209

Table 7
Summary of results for the interactions of age and gender on discipline events.a
Signicant (yes); not signicant (no)

Age: main
effect

Age gender:
interaction

Gender: main
effect

MannWhitney U test

Discipline event length


Total discipline duration
Maternal vocalization length during discipline
Maternal vocalization word count during discipline
Maternal vocalization tone
Use of explanation/reasoning
Use of affectionate terms
Infant negative behavior

No
Yes
Yes
No
Yes
Yes
No
No

Yes
No
Marginal
Yes
No
No
No
No

No
Marginal
Yes
Yes
No
No
Marginal
Yes

More affectionate terms to males


Males have more negative behavior

For specics of statistical analysis see Tables 16 and the text of Section 3.

response to it, measured in seconds). It was hypothesized that male infants would have shorter latencies than female infants,
and that latencies would decrease as infants grew older. Neither the interaction nor the main effects were signicant.
3.6. Summary of gender and age results
Because of the diversity of our test results we have produced a single table summary (Table 7) for signicant or nonsignicant results for tests of age, gender and agegender interactions. The details of our statistical analysis have been
presented in the previous section. The most surprising ndings are that, while male infants exhibit more negative behavior,
mothers appear to respond to males with more affectionate words, and use more words and talk for longer to boys during
discipline events.
4. Discussion
4.1. Overview
Little observational work has been done on discipline interactions between infants under the age of 12 months and their
mothers, and most existing work is cross-sectional. This study of 8 boys and 8 girls and their mothers, observed monthly from
5 to 12 months is a rst step toward lling this knowledge gap. Based on the literature, we hypothesized that male infants
would have higher rates of discipline events (due to both higher rates of misbehavior and mothers heightened tendency
to respond to male infants misbehaviors) and would be more emotionally reactive during discipline interactions; in turn,
we thought that mothers would use harsher forms of discipline toward their male infants. Instead, we found that although
male infants displayed more negative affect during discipline events, mothers used milder forms of discipline with male
than with female infants. During discipline events mothers vocalized more to male infants and used more words and more
affectionate terms with male than female infants.
One of the most surprising ndings was that membership in the high discipline group was not associated with any
maternal SES variables, but it was associated with lower father occupational status. The fathers of infants in the high
discipline group had signicantly lower occupation statuses than those in the low discipline group. While it would seem
more likely for mothers SES to affect the nature of motherinfant interactions than fathers, as fathers were not included
in our taped observations, ndings that fathers class status correlate with patterns of child management difculty are not
unprecedented (Belsky et al., 1996).
4.2. Child age and temperament
We found that as infants grow older, mothers subtly change their behavior during discipline events. When compared
with younger infants, mothers of older infants vocalized longer during discipline events, were more likely to explain the consequences of their infants misbehavior, and were more likely to use harsher, prohibitive tones. These three developmental
ndings (an increase in mothers vocalization length, use of explanations, and use of harsher tones over time) have not previously been documented in 512-month-old infants. However, they are not surprising. Regalado et al.s (2004) decision not
to ask mothers with infants under the age of 19 months about their use of explaining was based on the belief that mothers
of younger children would rarely use this discipline strategy. Our ndings show that the use of explanations increases as
children grow older during the rst year of life, but also that even mothers of young infants regularly utilize this technique.
In our study the coding of a discipline event required an infant misbehavior and a maternal response, both of which had
to meet certain criteria. Each of the mothers had slightly different standards for which behaviors were acceptable, which
behaviors could have negative or dangerous outcomes, etc. Some mothers were more likely to interpret seemingly benign
infant actions as misbehaviors. Still, our qualitative impressions suggest that the children with the highest rates of discipline
events also had the highest rates of clear-cut misbehavior. These impressions also accord with the positive relationship we
found between difcult infant temperament and more discipline events, suggesting that the childrens overall dispositions

R.E. Ahl et al. / Infant Behavior & Development 36 (2013) 199209

207

partially inuenced the number of discipline events. We found that the group of 7 children with the highest average number
of discipline events per observation had signicantly sadder and fussier moods at the ages of 6 and 9 months, and higher
overall temperamental difculty at 9 months, than the group with the lowest number of discipline events. Our ndings are
consistent with those of Kochanska et al. (2001), Socolar et al. (2005), Wissow (2002), Regalado et al. (2004), and Hendrix
(2004), all of whom found a relationship between temperament proles and parental discipline practices. Since there were
no signicant differences in temperament scores between the male and female children, our ndings of gender differences
cannot be attributed to gender differences in temperament.
4.3. Gender analyses
Male infants between 5 and 12 months had an average of 1.3 discipline events per 7-min observation, and female infants
had an average of .75. These rates are similar to those found by other studies, attesting to the validity of our methods and
results. Power and Chapieskis (1986) study of infants with a mean age of 13.7 months found a rate of one discipline event
every 9 min. Belsky et al.s (1996) study of male infants between 15 and 21 months found a rate of one discipline event
every 6 min.
We also found gender differences, some marginal, in the lengths and types of maternal and infant behaviors during
discipline events. In the 910 month age group, the average discipline event length for male infants (10.65 s) was signicantly
longer than the average length for female infants (6.07). Males had marginally, but not signicantly, longer total discipline
event durations than females. (Duration analyses relate to both the length of discipline events and the number of discipline
events.) The lack of signicant ndings for the latency code suggests that the greater amount of time male infants spend in
discipline events cannot be attributed to their mothers waiting longer before disciplining them; rather, the gender differences
are being driven by what happens once the mother starts to intervene. The length of the maternal intervention is a product
of both how long the mother chooses to issue discipline (through vocalization and/or through physical intervention) and
the childs reaction to her discipline. However, due to the generally low levels of observed overt infant deance, it seems
unlikely that infant deance was responsible. Instead, longer maternal interventions, irrespective of infant behavior, seemed
to lie at the heart of the difference. The length of maternal physical interventions was not measured in this study, but since
results from measures of maternal vocalization length showed that mothers of male infants used longer vocalizations, their
longer discipline behaviors cannot be attributed solely to longer physical interventions.
Mothers of male infants used signicantly longer vocalizations and more words than mothers of female infants during
discipline events, a pattern opposite to what we had hypothesized. A literature review on child genders effects on maternal
vocalization found that mothers talk more to female infants than male infants in general, in a variety of non-disciplinary
contexts (Fausto-Sterling, Garca Coll, & Lamarre, 2012). Thus, mothers of boys are using longer vocalizations during discipline
events despite having a general prole of shorter and fewer vocalizations in non-discipline situations.
Another reason why the nding of longer vocalizations to male infants is surprising is that previous research links longer
vocalizations to milder (Pffner & OLeary, 1989) and warmer (Grusec & Goodnow, 1994) discipline, and found that mothers
of male children are less likely to use these forms of discipline (Kochanska et al., 2003; Lytton & Romney, 1991). On the other
hand, longer vocalization and discipline event bouts could be associated with coercive discipline in older children (Belsky
et al., 1996; McFadyen-Ketchum et al., 1996); but we did not observe much coercive discipline in the present range of
observations. One could argue that longer vocalization is actually an indication of harsher discipline, but in this study, longer
vocalizations seemed to be associated with lengthier explanations. Pffner and OLeary (1989) found that longer, milder
verbal discipline was less effective for children between the ages of 18 and 31 months than shorter, harsher discipline,
but other researchers have found short vocalizations lacking in explanations to have detrimental long-term consequences
(Kochanska et al., 2003; Miner & Clarke-Stewart, 2008; van Zeijl et al., 2007). Due to the lack of consensus in the literature, it is
unclear as to whether mothers of boys use of longer vocalizations indicate either adaptive or problematic discipline practices.
The fact that there were no signicant differences in vocalization length or word count between mothers of male and
female infants during 1112 months could suggest that the differences seen at younger ages disappear at this age group.
Or, it could be attributed to specic data characteristics in this age group caused by the coding systems denitions. In the
1112 month age group, it became more common for short discipline events of the same type to happen in succession (for
instance, the mother removes a toy from the infants mouth, the infant puts the toy back in his or her mouth, and then this
process repeats). Each discipline event and accompanying vocalization would be short, but the overall vocalization duration
for the total observation would be long. Overall vocalization duration, however, was not measured in this study.
The nding of longer vocalizations to male infants does not by itself help us place mothers discipline of male and female
infants on a spectrum from harsh to mild. With the results of affectionate term usage analyses, a more complete picture
emerges. Mothers of male infants used more affectionate terms during their vocalizations than mothers of female infants.
This suggests that mothers of males may be using milder discipline than mothers of females. In summary, none of our analyses
of maternal discipline practices revealed that mothers of boys used harsher or more power-assertive discipline than mothers
of girls, while three of the analyses suggested that mothers of boys may be using milder discipline than mothers of girls.
For the infant behaviors, only one analysis yielded signicant results. Male infants were more likely to display negative
affect during discipline events than female infants. (The main effect for gender was marginally signicant using the mixed
between-within subjects ANOVA, and signicant using the MannWhitney U test.) The main effect for age was not signicant,
although it is worth noting that, for female infants, the percentage of discipline events in which negative affect was displayed

208

R.E. Ahl et al. / Infant Behavior & Development 36 (2013) 199209

declined from 11.6% during 58 months to 1.9% during 912 months. For male infants, the percentage only declined from
18.7% to 15.0%.
Because previous research found boys above the age of 2 to be more likely to respond to aversive events with aggression
than girls, we had hypothesized that males would display more negative affect than females (McFadyen-Ketchum et al.,
1996). We reasoned that a negative affect response could lead to harsher maternal discipline. The negative affect results
accord with the rst part of the hypothesis. However, the results on vocalization length and use of affectionate terms are at
odds with the second part. The results raise the possibility that, between the ages of 5 and 12 months, male infants greater
display of negative affect might cause their mothers to use milder disciplinary responses. According to this reasoning,
because male infants are more likely to cry, whimper, etc. during discipline events, their mothers use longer vocalizations
and more affectionate terms to prevent or shorten these displays of negative affect. It is easy to see how, as infants pass the
12-month mark, mothers responses could change. Maternal patience could wear out as the children approach the terrible
twos, when rates of misbehavior and externalizing behavior increase tremendously (Arnold, OLeary, Wolff, & Acker, 1993;
Miner & Clarke-Stewart, 2008). Once rates of child misbehavior and negative affect reach a certain threshold, mothers might
respond with harsher, rather than milder, discipline techniques. Such a dynamic reversal of dyadic behavior ought to be
detectable as a phase transition between a stable attractor for the motherson dyad during year one to a different (and more
negative) attractor appearing during at some point after 12 months of development (Hollenstein, 2007).
4.4. Limitations, contributions, and further study
Our ndings on gender differences could stem from elements of the discipline interactions that this study did not sufciently examine, such as infant misbehavior type. Inferential statistics were not performed on the infant misbehavior type
codes due to the relatively small numbers of misbehaviors in each category. An examination of the descriptive statistics
suggests that some gender and age-based misbehavior type differences might be important topics for future studies. For
instance, boys seemed to have a higher percentage of harm to mother misbehaviors across all ages.
The type of physical interventions that took place during discipline events might also be important. Physical intervention
type (i.e. moving the infant, removing object from infant, etc.) was initially part of the Coding System, but results from this
code were not analyzed because we found the physical intervention code to be too dependent on infant misbehavior type
to be meaningfully examined without the infant misbehavior type information. (For instance, toy removal was only coded if
the infant was misusing an object or playing with a forbidden object, whereas moving the infants body was usually coded if
the infant was in a forbidden location.) Possibly, mothers of males used more forceful physical interventions than mothers
of females, and the use of physical interventions contributed to male infants higher levels of negative affect.
Finally, our ndings should not be over-generalized. Participating dyads were Rhode Islanders from white, predominantly
middle-class backgrounds. Due to the substantial body of literature demonstrating cross-cultural and class differences in
discipline practices (Bernstein, Harris, Long, Iida, & Hans, 2005), it cannot be assumed that our results would hold true for all
segments of the American population or for other national groupings. Furthermore, the original videotaping was done about
20 years ago, and it is entirely possible that disciplinary practices have changed in that time period. Culturally and historically,
parental practices are a moving target. Nevertheless, we have outlined a method that could be used in comparative studies
and have shown the importance of studying discipline in infancy.
4.5. Conclusion
By employing a longitudinal, naturalistic, observation-based approach to the study of young infants we have contributed
new ndings to the body of literature on discipline. We have found signicant age-based and gender-based differences in
discipline patterns. The most notable results were that during discipline interactions, male infants displayed signicantly
more negative emotions than female infants and mothers of male infants used signicantly longer vocalizations, and more
affectionate terms, than mothers of female infants. The overall patterns seemed to suggest higher child difculty in male
infants, and milder discipline usage amongst mothers of male infants.
Using a methodology that investigates the details of dyadic discipline interactions longitudinally and cross modally allows
us to address the sources of gender differences in these interactions. None of the signicant results of this study would have
been discovered if surveys had been used instead of observations. Dynamic systems theory argues that small changes, or
pushes in one direction, become magnied with the passage of time. Small increases in male infant negative affect, for
instance, could lead to excessively harsh or lax maternal discipline practices (Lorber & Slep, 2005), causing increasingly
problematic discipline patterns as boys grow older. The study of how mothers discipline their infants has tremendous
potential to illuminate the causes of later gender differences in externalizing and aggressive behavior, as well as potential
avenues for interventions. It may be easier to encourage healthy discipline practices in mothers when their infants are young,
and the rates of discipline events are low.
Acknowledgement
We would like to thank and acknowledge the assistance of Ms. Christelle Ngnoumen, Dr. Deborah Schooler and Dr. Jihyun
Sung.

R.E. Ahl et al. / Infant Behavior & Development 36 (2013) 199209

209

Appendix A. Supplementary data


Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/
j.infbeh.2013.01.005.
References
Arnold, D. S., OLeary, S. G., Wolff, L. S., & Acker, M. M. (1993). The Parenting Scale: A measure of dysfunctional parenting in discipline situations. Psychological
Assessment: 5., (2), 137144.
Belsky, J., Woodworth, S., & Crnic, K. (1996). Trouble in the second year: Three questions about family interaction. Child Development: 67., (2), 556578.
Bernstein, V. J., Harris, E. J., Long, C. W., Iida, E., & Hans, S. L. (2005). Issues in the multi-cultural assessment of parentchild interaction: An exploratory
study from the starting early starting smart collaboration. Journal of Applied Developmental Psychology: 26., (3), 241275.
Casas, J. F., Weigel, S. M., Crick, N. R., Ostrov, J. M., Woods, K. E., Yeh, E. A. J., et al. (2006). Early parenting and childrens relational and physical aggression
in the preschool and home contexts. Journal of Applied Developmental Psychology: 27., (3), 209227.
Fausto-Sterling, A., Garca Coll, C., & Lamarre, M. (2012). Sexing the baby: Part 1 What do we really know about sex differentiation in the rst year of life?
Social Science and Medicine: 74., (11), 16841692.
Granic, I., Hollenstein, T., Dishion, T. J., & Patterson, G. R. (2003). Longitudinal analysis of exibility and reorganization in early adolescence: A dynamic
systems study of family interactions. Developmental Psychology: 39., (3), 606617.
Grusec, J. E., & Goodnow, J. J. (1994). Impact of parental discipline methods on the childs internalization of values: A reconceptualization of current points
of view. Developmental Psychology: 30., (1), 419.
Hendrix, R. R. (2004). Emergence of parental prohibition in response to self-produced locomotion and its accompanying emotional changes. Lincoln: University
of Nebraska.
Hollenstein, T. (2007). State Space Grids: Analyzing dynamics across development. International Journal of Behavioral Development: 31., 384396.
Hollenstein, T., Granic, I., Stoolmiller, M., & Snyder, J. (2004). Rigidity in parentchild interactions and the development of externalizing and internalizing
behavior in early childhood. Journal of Abnormal Child Psychology: 32., (6), 595607.
Kawabata, Y., Alink, L. R. A., Tseng, W.-L., van Ijzendoorn, M. H., & Crick, N. R. (2011). Maternal and paternal parenting styles associated with relational
aggression in children and adolescents: A conceptual analysis and meta-analytic review. Developmental Review: 31., (4), 240278.
Kochanska, G. (2001). Emotional development in children with different attachment histories: The rst three years. Child Development: 72., (2), 474490.
Kochanska, G., Aksan, N., & Nichols, K. E. (2003). Maternal power assertion in discipline and moral discourse contexts: Commonalities, differences, and
implications for childrens moral conduct and cognition. Developmental Psychology: 39., (6), 949963.
Kochanska, G., Coy, K. C., & Murray, K. T. (2001). The development of self-regulation in the rst four years of life. Child Development: 72., (4), 10911111.
Kochanska, G., Coy, K. C., Tjebkes, T. L., & Husarek, S. J. (1998). Individual differences in emotionality in infancy. Child Development: 69., (2), 375390.
Kochanska, G., Kuczynski, L., & Radke-Yarrow, M. (1989). Correspondence between mothers self-reported and observed child-rearing practices. Child
Development: 60., (1), 5663.
Kochanska, G., Tjebkes, T. L., & Forman, D. R. (1998). Childrens emerging regulation of conduct: Restraint, compliance, and internalization from infancy to
the second year. Child Development: 69., (5), 13781389.
Laumann, E. O., Michael, R. T., & Gagnon, J. H. (1994). A political history of the national sex survey of adults. Family Planning Perspectives: 26., (1), 3438.
Lorber, M. F., & Slep, A. M. (2005). Mothers emotion dynamics and their relations with harsh and lax discipline: Microsocial time series analyses. Journal of
Clinical Child and Adolescent Psychology: The Ofcial Journal for the Society of Clinical Child and Adolescent Psychology, American Psychological Association,
Division 53: 34., (3), 559568.
Lytton, H., & Romney, D. M. (1991). Parents differential socialization of boys and girls: A meta-analysis. Psychological Bulletin: 109., (2), 267296.
McFadyen-Ketchum, S. A., Bates, J. E., Dodge, K. A., & Pettit, G. S. (1996). Patterns of change in early childhood aggressivedisruptive behavior: Gender
differences in predictions from early coercive and affectionate motherchild interactions. Child Development: 67., (5), 24172433.
Miner, J. L., & Clarke-Stewart, K. A. (2008). Trajectories of externalizing behavior from age 2 to age 9: Relations with gender, temperament, ethnicity,
parenting, and rater. Developmental Psychology: 44., (3), 771786.
Nelson, D. A., Hart, C. H., Yang, C., Olsen, J. A., & Jin, S. (2006). Aversive parenting in China: Associations with child physical and relational aggression. Child
Development: 77., (3), 554572.
Ostrov, J. M., & Keating, C. F. (2004). Gender differences in preschool aggression during free play and structured interactions: An observational study. Social
Development: 13., (2), 255277.
Pallant, J. (2007). SPSS Survival Manual. England: Open University Press.
Pffner, L. J., & OLeary, S. G. (1989). Effects of maternal discipline and nurturance on toddlers behavior and affect. Journal of Abnormal Child Psychology: 17.,
(5), 527540.
Power, T. G., & Chapieski, M. L. (1986). Childrearing and impulse control in toddlers: A naturalistic investigation. Developmental Psychology: 22., (2), 271275.
Ravn, I. H., Smith, L., Lindemann, R., Smeby, N. A., Kyno, N. M., Bunch, E. H., et al. (2011). Effect of early intervention on social interaction between mothers
and preterm infants at 12 months of age: A randomized controlled trial. Infant Behavior and Development: 34., (2), 215225.
Ravn, I. H., Smith, L., Smeby, N. A., Kynoe, N. M., Sandvik, L., Bunch, E. H., et al. (2012). Effects of early motherinfant intervention on outcomes in mothers
and moderately and late preterm infants at age 1 year: A randomized controlled trial. Infant Behavior and Development: 35., (1), 3647.
Regalado, M., Sareen, H., Inkelas, M., Wissow, L. S., & Halfon, N. (2004). Parents discipline of young children: Results from the National Survey of Early
Childhood Health. Pediatrics: 113., (6 Suppl.), 19521958.
Seifer, R., Sameroff, A. J., Barrett, L. C., & Krafchuk, E. (1994). Infant temperament measured by multiple observations, mother report. Child Development:
65., 14781490.
Socolar, R. R., Savage, E., Keyes-Elstein, L., & Evans, H. (2005). Factors that affect parental disciplinary practices of children aged 12 to 19 months. Southern
Medical Journal: 98., (12), 11811191.
Straus, M. A., Hamby, S. L., Finkelhor, D., Moore, D. W., & Runyan, D. (1998). Identication of child maltreatment with the Parent-Child Conict Tactics
Scales: Development and psychometric data for a national sample of American parents. Child Abuse and Neglect: 22., (4), 249270.
van Zeijl, J., Mesman, J., Stolk, M. N., Alink, L. R., van Ijzendoorn, M. H., Bakermans-Kranenburg, M. J., et al. (2007). Differential susceptibility to discipline:
The moderating effect of child temperament on the association between maternal discipline and early childhood externalizing problems. Journal of
Family Psychology: JFP: Journal of the Division of Family Psychology of the American Psychological Association: 21., (4), 626636.
Wissow, L. S. (2002). Child discipline in the rst three years of life. In N. Halfon, K. T. McLearn, & M. A. Schuster (Eds.), Child rearing in America: Challenges
facing parents with young (pp. 146177). Cambridge: Cambridge University Press.

Vous aimerez peut-être aussi