Journal of Child Psychology and Psychiatry 46:1 (2005), pp 19–34
doi: 10.1111/j.1469-7610.2004.00348.x
Pathways to conscience: early mother–child
mutually responsive orientation and children’s
moral emotion, conduct, and cognition
Grazyna Kochanska, David R. Forman, Nazan Aksan, and Stephen B. Dunbar
University of Iowa, USA
Background: Associations between early mother–child mutually responsive orientation (MRO) and
children’s conscience have been previously established, but the mechanisms accounting for those links
are not understood. We examined three such mediational mechanisms: (a) the child’s enhanced
enjoyment of interactions with the mother, (b) increased committed, self-regulated compliance with the
mother, and/or (c) a decreased need for maternal use of power assertion. Children’s conscience was
seen as a complex system encompassing moral emotion (guilt), conduct, and cognition. Methods: In a
longitudinal design, MRO was observed in mothers’ and children’s multiple naturalistic interactions
at 9, 14, and 22 months. The mediators were observed at 33 months. Children’s conscience was
observed at 45 months (moral emotion) and at 56 months (moral conduct and cognition). Results: The
mediating paths were different for the three components of conscience. MRO had a direct, unmediated
effect on moral emotion. MRO influenced moral conduct through two mediational paths: by promoting
the child’s enjoyment of interactions with the mother and by enhancing committed compliance. MRO
influenced moral cognition by promoting the child’s enjoyment of mother–child interactions. Maternal
power assertion did not mediate the relation between MRO and conscience once the influence of
the other mediators was considered. Conclusions: The impact of the early mother–child relationship on future conscience appears to be a complex process that progresses along distinct
paths. Keywords: Conscience, mother–child relationship, developmental pathways. Abbreviation: MRO: mutually responsive orientation.
How children gradually internalize rules and standards of conduct, and how individual consciences
collectively support the inter-generational transmission of values, have been compelling questions from
the beginning of recorded history (Grusec, 1997). An
impaired early conscience is a risk factor for, and a
core aspect of, future conduct problems, antisocial
development, and psychopathy (Blair, 1997; Frick
et al., 2003; Lykken, 1995). Conversely, early capacity for remorse, engaging in conduct compatible
with rules, and an understanding of right and wrong
are markers of successful adaptation. Recently, we
have returned to the study of conscience, with a new
appreciation for the early years (Eisenberg & Fabes,
1998; Emde, Biringen, Clyman, & Oppenheim,
1991; Grusec & Goodnow, 1994; Kochanska, 1993;
Radke-Yarrow, Zahn-Waxler, & Chapman, 1983;
Thompson, 1998).
Parental influence on children’s conscience
development remains controversial (Bugental &
Goodnow, 1998; Collins, Maccoby, Steinberg,
Heatherington, & Bornstein, 2000; Eisenberg, 1998;
Maccoby, 1992; Maccoby & Martin, 1983). The role
of parents has been revisited owing to the blossoming ‘science of relationships’ (Hinde, 1979; Hartup,
1989). Relationships endure over time, are built
upon the history of interactions, and involve strong,
frequent, and diverse connections (Maccoby, 1999).
They are critical developmental contexts, which
afford a productive vantage point from which to
explore early socialization (Collins & Laursen, 1999;
Maccoby, 1992; Reis, Collins, & Berscheid, 2000;
Thompson, 1998, 2000).
As one approach consistent with that perspective,
we have proposed a concept of mutually responsive orientation (MRO) (Kochanska, 1997, 2002a;
Kochanska, Forman, & Coy, 1999). This approach
has evolved in the context of several converging traditions, all sharing a common premise: Relationships characterized by mutual responsiveness,
positivity, and harmony promote partners’ eager,
willing stance toward each other’s influence.
Maccoby (1983, 1999; Maccoby & Martin, 1983)
viewed mutuality and reciprocity as the ‘positive
socialization forces’ that promote child cooperation
with the parents. Attachment scholars believe that
secure children, who presumably have had responsive care, are cooperative and well internalized (Belsky,
1999; Bretherton, Golby, & Cho, 1997; Londerville &
Main, 1981; Matas, Arend, & Sroufe, 1978; Stayton,
Hogan, & Ainsworth, 1971; van IJzendoorn, 1997).
Clark (1984) described a ‘communal relationship’
between adult partners who are invested in each
other’s welfare, empathic, responsive to one another,
and share a sense of mutual reciprocity. Integrating
those literatures, we proposed that MRO is a positive,
close, mutually binding, cooperative relationship
that emerges in some parent–child dyads, and promotes many positive socialization outcomes, including children’s conscience. We also proposed that
Association for Child Psychology and Psychiatry, 2004.
Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
20
Grazyna Kochanska et al.
MRO has two components: mutual responsiveness
and shared affective positivity.
Mutual responsiveness in relationships has been
seen as promoting trust in the partner, security,
mutual bond, sense of agency and efficacy, empathy,
and expectations of reciprocity. Parental responsiveness has been linked to child cooperation, and its
lack to impaired socialization (Bryant & Crockenberg, 1980; Lytton, 1980; Martin, 1981; Parpal &
Maccoby, 1985; Rocissano, Slade, & Lynch, 1987;
Shaw, Keenan, & Vondra, 1994; Westerman, 1990).
Shared affective positivity – mutually experienced
pleasurable, smoothly flowing activities infused with
positive emotion in both partners – has been studied
less than responsiveness. Its importance, however,
has been stressed by many (Dix, 1991; Emde et al.,
1991; Lay, Waters, & Park, 1989; Maccoby, 1983;
Maccoby & Martin, 1983; Radke-Yarrow, Richters, &
Wilson, 1988).
In two large longitudinal studies, we replicated
robust findings of positive implications of MRO for
conscience, across a striking range of the designs,
ages, and outcomes: empathy, guilt, internalized
conduct alone and with peers, reluctance to transgress, and maternal reports (Kochanska, 1997,
1998; Kochanska et al., 1999; Kochanska & Murray,
2000). The links were concurrent (toddler, preschool
age) and longitudinal (toddler to preschool age,
toddler and preschool age to early school age, infancy to toddler age), and supported by others
(Bugental, 2000; Deater-Deckard & O’Connor, 2000;
Deater-Deckard & Petrill, 2004; De Wolff & van IJzendoorn, 1997; Harrist & Waugh, 2002; Laible &
Thompson, 2000; Maccoby, 1999; Thompson,
1998).
We do not yet know, however, what mechanism or
mechanisms account for those established empirical
links between MRO and children’s conscience. To
identify such potential mechanisms is a necessary
next step, a goal consistent with the recent agenda in
social-emotional development (Eisenberg, 1998).
This goal is made challenging by the complex
nature of conscience, which, even at a young age,
encompasses emotional, behavioral, and cognitive
components. Although related, those components
may not form a unitary, global predisposition, as
first pointed out by Hartshorne and May (1928–
1930), and revisited since (Burton, 1963; Sears,
Rau, & Alpert, 1965). Thus, we asked if the mediating paths linking early MRO and later conscience
may be relatively specific to the component of conscience being studied.
To examine the paths linking early MRO with
children’s future conscience is the objective of this
article. We propose three mediators of those links:
MRO may exert its influence by (a) fostering child
enjoyment of the interaction with the parent, (b)
promoting committed compliance, a nascent form of
internalization, and/or (c) lessening the need for
parental power assertion.
The first posited, and tested, mechanism involves
the child’s enjoyment of interactions with the parent.
In mutually responsive relationships, children come
to feel happy and content when interacting with
parents, perhaps due to the prevailing positive
ambience, or because responsive parents prevent
them from experiencing many negative emotions.
Positive emotions have been consistently and
broadly linked to multiple forms of prosocial behavior in social psychology (Carlson, Charlin, & Miller,
1988; Isen, 1999). Developmentalists have also
acknowledged broad benefits of positive mood for
socialization, although they focus on more enduring
positive emotions and developmental processes, rather than on situational mood (Emde et al., 1991;
Eisenberg & Fabes, 1998; Lay et al., 1989; Maccoby,
1983; Waters, Kondo-Ikemura, Posada, & Richters,
1990). We examined if such a link could be empirically shown, and further, whether such enjoyment
indeed mediated the links between early MRO and
future moral conduct and cognition.
We did not, however, expect those positive feelings
to promote children’s tendency to experience the
moral emotion of guilt. Unlike prosocial cognition
and conduct, guilt tends to be associated with negative rather than positive emotionality (Kochanska,
DeVet, Goldman, Murray, & Putnam, 1994).
The second proposed mediating link involves
committed compliance, eager, wholehearted, and
self-sustained form of compliance (Kochanska &
Aksan, 1995; Kochanska, Aksan, & Koenig, 1995;
Kochanska, Coy, & Murray, 2001). Committed
compliance, first shown with the mother present,
over time leads to fledgling, genuine internalization
of maternal rules – rule-compatible conduct in the
absence of control and surveillance, originating
entirely within the child.
Our earlier work has shown both the correlations
between mother–child MRO and child committed
compliance, and between committed compliance and
emerging conscience. Those links were concurrent
and longitudinal, across a broad range of ages (from
14 to 48 months) and measures: rule-compatible
behavior in many settings, and multiple maternal
ratings (Kochanska, 2002b; Kochanska & Aksan,
1995; Kochanska et al., 1995; Kochanska, Tjebkes,
& Forman, 1998; Kochanska et al., 2001). Again,
however, we have never examined the mediational
character of those empirical associations. In this
study, we tested whether committed compliance
serves as a mediator linking MRO with children’s
moral emotion, moral conduct, and moral cognition.
The third mediating mechanism involves maternal
use of power. High MRO may reduce the need for
parents to rely on forceful discipline. Indeed, we have
found that in mutually responsive dyads, mothers
used less power in control encounters. The links
were between MRO at toddler and preschool age and
power at toddler and preschool age, concurrent and
longitudinal, and between MRO at 9 and 14 months
Mutually responsive orientation
and maternal power at 22 months (Kochanska,
1997; Kochanska et al., 1999), across observed and
self-reported power measures.
Parental power, in turn, has been often seen as
undermining children’s internalization (Grusec &
Goodnow, 1994; Hoffman, 1983; Kochanska,
Padavich, & Koenig, 1996; Lepper, 1981). Power
assertion may engender external attributions for
compliance (Lepper, 1981), a ‘threat to autonomy’
(Grusec & Goodnow, 1994), and resentment and
anger toward the parents, and the consequent rejection of the parent’s values. It may lead the child to
store parents’ messages in episodic rather than
semantic memory, and thus impair internalization
(Hoffman, 1983). Thus, we tested if the decreased
power assertion may indeed mediate the link
between MRO and conscience.
In sum, we examined three non-mutually exclusive models explaining the links between MRO and
conscience. We propose that MRO promotes conscience by: (a) fostering the child’s enjoyment of interactions with the parent, (b) promoting the child’s
committed compliance, and/or (c) diminishing the
need for parental power assertion.
To strengthen our ability to address the causal
hypotheses, we adopted a developmental longitudinal approach. The measures of MRO came from
behavioral observations of extensive naturalistic
mother–child interactions at 9, 14, and 22 months.
The three mediators (child enjoyment of interactions,
committed compliance, and maternal use of power)
were assessed at 33 months, also using observations
in multiple contexts. The measures of children’s
conscience included their moral emotion, guilt, when
they believed they had committed a transgression (at
45 months), moral conduct when they were alone in
situations that called for resisting strong temptations to break rules (at 56 months), and cognitive
responses to hypothetical moral dilemmas (also at
56 months).
Mediation was examined for each conscience
measure and for each mediational path separately
(Baron & Kenny, 1986), in sequential regressionbased tests of predictors and mediating influences.
We also tested all three mediators in one comprehensive regression. In all analyses, to control for the
systematic variation in both the mediators and outcomes (Collins et al., 2000), we used their earlier
measures, concurrent with MRO, as covariates,
when available (the composites of the 14- and 22month scores for all three mediators and for child
moral conduct, and a 22-month score, available for
moral emotion).
Method
Participants
Mothers of normally developing infants in intact families in Iowa replied to ads in local papers, libraries,
21
day care and health care centers, etc. The families
were mostly White (mothers, 97%, fathers 92%), but
ranged broadly in income and education (details are
in Kochanska, Coy, Tjebkes, & Husarek, 1998a). We
report data from 9 months, M ¼ 8.94, SD ¼ .63,
N ¼ 112 (56 girls), 14 months, M ¼ 13.65, SD ¼ .74,
N ¼ 108 (53 girls), 22 months, M ¼ 22.30, SD ¼ .55,
N ¼ 106 (53 girls), 33 months, M ¼ 32.80, SD ¼ .53,
N ¼ 104 (52 girls), and 56 months, M ¼ 56.53,
SD ¼ 1.09, N ¼ 74 (41 girls). The study had been
originally planned through 45 months; when additional funds became available, some participants
had left the area or become too busy. The families
who did and did not return did not differ on any
relevant variable, based on t-tests.
Overview
At 9 months, we conducted a 2–2½ hr home session;
at 14 months, a 2–2½ hr laboratory session. At 22
and 33 months each, there were two 3–4 hr laboratory sessions, and at 56 months – one 4 hr laboratory session. Each session involved the mother and
the child, and was conducted by a young woman
(‘E’). The laboratory has a naturalistically furnished
living room and a sparsely furnished play room. The
sessions were videotaped for later coding. Independent teams coded all data sets. Reliability was
based on at least 15% of cases. The coders realigned
routinely to prevent observer drift. To obtain robust
constructs, we aggregated the data, following standardization, at multiple levels of measurement
(Rushton, Brainerd, & Pressley, 1983).
Predictor measures: mutually responsive
orientation (9, 14, 22 months)
MRO encompassed two components: maternal
responsiveness and shared positivity. The coded
contexts encompassed naturalistic mother–child
interactions in many daily situations, such as free
time, meal preparation, meal time, routine caregiving (giving the child a bath, changing his/her
clothes), play time, etc. The approximate total coded
times were: 60 min at 9 months, 40 min at
14 months, and 90 min at 22 months. The coding,
reliability, data reduction, and coherence for the 9and 14-month data are in Kochanska et al. (1999).
Here, we also include 22-month data. Because the
final measures were analogous with the past publications, their description is brief. Reliability at
22 months is described.
Maternal responsiveness
Two coding systems were used: microscopic and
macroscopic coding (see Kochanska, 1997). They
were coded independently, and their final scores
were ultimately aggregated.
22
Grazyna Kochanska et al.
Microscopic coding. A combination of a time-sampled and event-triggered approach entailed two
passes through a videotape, using 60-s intervals.
During the first pass, the coders decided whether or
not the child directed a signal toward the mother or
was in a state requiring maternal response (reliability, kappa .87), and they identified all such signals
(for example, child distress or bid; kappa .83). In the
absence of any signal, one of several global codes
was given (kappa .90). Only one global code was
considered here (continuous harmonious exchange),
used in the measure of shared positivity, described
below.
During the second pass, the coders located each
child signal, and evaluated the mother’s response to
it as poor, fair, good, or exceptional (kappas .68–.75),
considering multiple aspects of responsiveness:
sensitivity, promptness, engagement, acceptance,
cooperation, emotional availability, ability to follow
child lead and focus of attention, and to adjust
stimulation to child state.
Data reduction. We tallied all the instances when
the mother responded poorly, fairly, well, or exceptionally to the signals in each category. Each tally
was divided by the total number of the signals in that
category (e.g., for child distress category, we computed the proportions of all child distress signals to
which the mother responded poorly, fairly, well, or
exceptionally). Next, we computed four broader
composite scores which represented poor, fair, good,
and exceptional response patterns (respectively, the
averages of poor, fair, good, and exceptional response across all child signals).
Finally, we created an overall microscopic weighted
score by weighing the composite of poor responses by
)2, that of fair responses by )1, that of good responses by +1, and that of exceptional responses by
+2, and by summing these scores (the means at 9,
14, and 22 months, respectively: ).38, .18, ).16).
Macroscopic coding (ratings). Each interactive
context (e.g., play, snack) was rated using the three
classic 9-point scales (Ainsworth, Bell, & Stayton,
1971): sensitivity–insensitivity (awareness of and
attunement to the child’s signals), acceptance–
rejection (warmth, interest in and enjoyment of the
child), and cooperation—interference (respect for the
child’s autonomy). Reliability, kappas, ranged from
.65 to .83. For each context, all the scales correlated
(r’s above .72), and were averaged into one score,
then across contexts (alpha .88 for session 1 and .90
for session 2 at 22 months) into an overall macroscopic score (the means at 9, 14, and 22 months:
5.16, 4.89, 6.08).
Overall history of maternal responsiveness, 9 to
22 months. Despite having been produced by separate coding teams, the overall microscopic weighted
score and the overall macroscopic score correlated at
all ages (from .33 to .56, all p’s < .001); they were
standardized and averaged into an overall maternal
responsiveness score at each age (9, 14, 22 months).
Shared positivity
Mother–child shared positivity incorporated mother–
child shared positive affect scores and the continuous harmonious exchange global scores from the
responsiveness coding, described above.
Coding and data reduction. Shared positive affect
was coded by an independent team during the same
mother–child naturalistic interactions as was
maternal responsiveness. The coding segments were
30 s long. For both the mother and for the child, one
or more of the discrete negative affects (not used
here) or discrete positive affects (affection and joy)
was coded. If there was no ‘full-blown’ emotion,
neutral positive mood or neutral negative mood was
coded. Kappas were .68 to .79 for the mother, and
.63 to .80 for the child. We tallied all the segments
where both the mother and child displayed positive
affect or neutral positive mood and neither displayed
negative affect or neutral negative mood, and divided
those tallies by the number of coded segments (the
means at 9, 14, and 22 months, respectively, .59,
.81, .79).
The continuous harmonious exchange was coded
within the microscopic responsiveness system as a
global code (a segment when the child did not direct
any signals toward the mother, and the interaction
appeared harmonious and smooth). Those codes
were tallied and divided by the number of coded
segments (the means at 9, 14, and 22 months,
respectively, .23, .41, .12).
Those two scores correlated: at 9 months,
r (112) ¼ .55, at 14 months, r (108) ¼ .46, and at
22 months, r (106) ¼ .53, all p’s < .001. Thus, they
were standardized and aggregated into a shared
positivity score at each age.
Overall MRO score, 9–22 months
In the final step, at each age, the overall maternal
responsiveness score and the shared positivity
scores, which correlated (r’s ranged from .47 to .55,
p’s < .001), were aggregated into the final MRO score.
Next, these MRO scores, which were longitudinally
stable (r’s ranged from .34 to .50, all p’s < .001) were
aggregated into one score of the overall MRO score,
reflecting the history of MRO from 9 to 22 months.
Mediators at 33 months: children’s enjoyment of
interactions with mothers, children’s committed
compliance, and mothers’ power assertion
Children’s enjoyment of interactions with mothers.
Coding and data reduction. Working with the affect
codes for each child (affect was coded in a parallel
Mutually responsive orientation
manner to that described above, for about 60 min of
mother–child interaction across two sessions,
kappa ¼ .89), we weighted the instances of the
child’s positive affect that were pervasive or intense
(longer than half of a segment, and/or a strong
emotion) by 3, the discrete positive affect codes by 2,
and neutral positive affect codes by 1, added those
figures, and divided by the number of coded segments to create the score of the child’s enjoyment of
the interactions with the mother.
Children’s committed compliance. Coding. Committed compliance was coded in the laboratory,
during 68 minutes of mother–child interaction in
contexts such as snack, play, and free time. There
was also a low shelf with attractive toys and other
objects, designated as off-limits for the child (for
details, see Kochanska et al., 2001). The mother
issued this prohibition at the outset, and enforced it
throughout each session. Every episode when the
child was involved with the prohibited toys was
coded, for every 30-s segment, until the child reoriented.
In each segment, the predominant behavior was
coded, using several codes. Here, we use focus only
on committed compliance, coded when the child
looked but did not attempt to touch; verbalized the
prohibition (e.g., ‘no-no toys’, ‘we don’t touch’); or
turned away spontaneously. Restraint appeared to
originate within the child, not due to maternal
immediate control (Kochanska et al., 2001). Reliability for identifying the onset of an episode ranged
from 87% to 96%; kappas for the behavioral codes
were .65 to .77.
Data reduction. Tallied codes were divided by the
number of coded segments, which were variable,
because they depended on how many times the child
directed attention to the prohibited toys, thus triggering the beginning of a coded episode (for session
1, M ¼ 14.12, SD ¼ 6.96, and for session 2,
M ¼ 9.10, SD ¼ 5.35). Thus created relative committed compliance scores correlated across the sessions, r (103) ¼ .56, p < .001, and were averaged
across them, to create an overall score of committed
compliance.
Mothers’ power assertion. Coding. The mother’s
power assertion was coded in the discipline contexts
described above. For each 30-s segment when the
child was involved with the prohibited toys, the
coders assigned one global rating of maternal control
style (kappas .71–.85) and they coded all physical
interventions the mother used in the segment (.69–
.77). The global ratings that reflected power-assertive discipline included: assertive control (mother
controls in decisive, firm manner) and forceful control
(uses power, threats, negative or angry control). The
power-assertive physical interventions included
assertive physical control (holds child firmly, moves
23
him/her decisively, takes away a toy, etc.) and
forceful physical control (shakes, slaps, spanks,
handles roughly, uses frightening, angry gestures).
Data reduction. For each prohibition context, we
tallied all instances of the power-assertive codes,
divided each tally by the number of segments, averaged across the two sessions, standardized, and
created a power-assertive global and physical scores.
Those correlated, r(104) ¼ .43, p < .001, and were
aggregated into one score of maternal power assertion.
Outcome measures: children’s conscience
Moral emotion of guilt (45 months). Procedure. All
empirical details are in (Kochanska, Gross, Lin, and
Nichols 2002). In two 3-min ‘mishap’ paradigms, the
child was led to believe that he or she had transgressed (damaged E’s valued objects – a stuffed cat
and a toy boat).
Coding. Two codes were applied to child emotional
expression for every 5-s segment (average of 66).
They included: avoiding gaze (look away, down, or
askance; marked as brief or long), and bodily tension
(squirm, back away, hang down head, hunched
shoulders, hug self, cover face with hands). Three
codes were applied to longer epochs in the mishap
paradigm (for example, 60 s after the occurrence of
the mishap): overall distress response, from 1 (child
not distressed/affected by the mishap), to 4 (child
strongly distressed/affected, freezes, cries, uncomfortable, uneasy); negative affect and positive affect
(further qualified as strong, if applicable). Reliabilities (kappas) were: for gaze avoidance, .88 and .91;
for bodily signs of tension, .77 and .62; for overall
distress response, .61 and .63; for affect, .77 and
.60.
Data reduction. For each mishap, we tallied all
instances of each 5-s code (having weighted long
gaze avoidance), producing the scores of avoiding
gaze and bodily tension. We tallied the epoch-based
codes (having weighted strong affects), to produce
an overall distress score, a negative affect score,
and a positive affect score for each mishap. Those
five components of guilt cohered across the two
mishaps (mean r ¼ .48), and were aggregated.
Finally, we created a composite guilt score at
45 months by standardizing and averaging those
scores, having reversed the positive affect score;
alpha was .83.
Moral conduct (56 months). Internalization while
alone with prohibited toys. Procedure. The procedural and coding details are in Kochanska et al.
(2001). At the end of the session, during which the
mother enforced the prohibition regarding the offlimits toys, we observed child internalization. The
24
Grazyna Kochanska et al.
mother repeated the prohibition, and left to the
adjoining room, leaving the child alone. To draw the
child’s attention to the toys, she also asked the child
to do a dull sorting task, set up directly in front of
the shelf. The child was alone for 1 min; then, an
unfamiliar female entered, played with the most
attractive toys (1 min), and left; child was again
alone for 6 min.
Coding. The child’s behavior was coded for each of
96 5-s segments as: looking at toys without touching, engaged in other activity (e.g., snacking), sorting, touching toys gently, self-correcting (beginning
to touch and terminating the attempt spontaneously), and deviating (playing with the toys).
Latency to deviate was also coded. Reliability, kappa,
was .91.
Data reduction. The relative scores for each of the
coded behaviors (divided by the number of segments), and latency to deviate were submitted to
Principal Components Analysis. The first factor
produced by the PCA expressed child internalization:
low deviation ().94), long latency to deviate (.92),
sorting activity (.53), low gentle touch ().37), and
looking without touching (.26). Eigenvalue was 2.83,
accounting for 40% of variance. The score on this
factor reflected internalization while alone with prohibited toys.1
Internalization while playing the ‘cheating game’.
Procedure. The child played a game that involved
throwing velcro balls at a target on the wall for 3 min,
while left alone (Kochanska & Murray, 2000). E explained the rules and asked not to cheat, during a
friendly, but serious talk about the meaning of
cheating, honesty, etc. The rules made the game
impossible to win (e.g., child was asked to face away
from the target or stay within a confined space far
away). The child was to receive a prize for each ball
that stuck to the target. Upon return, E ‘discovered’
she had used ‘the wrong rules’ and let the child play
the game again; every child won a prize.
Coding and data reduction. Child behavior
was coded for every 3 s for both ‘legal’ and ‘illegal’
behaviors, such as facing away from the target,
or facing the target while throwing; remaining
inside the specified space or being outside of it.
Latencies to the first instances of the illegal behaviors were also coded. Reliability, kappa, was .90.
Regarding the latencies to the first occurrence of a
behavior, 93% of coders’ judgments were within 1 s.
The averaged latencies, legal behaviors, and
(reversed) illegal behaviors, all standardized, were
1
This approach has yielded a similar factor score repeatedly in
the past, across several samples and ages (e.g., Kochanska
et al. 2001).
aggregated into a score of internalization in the
cheating game.
Overall moral conduct. The two measures of internalized behavior in the absence of surveillance (with
the prohibited toys and in the ‘cheating game’) correlated, r(74) ¼ .49, p < .001, and thus were aggregated into one moral conduct score at 56 months.
Moral cognition (56 months). Procedure. The procedure was similar to that in an earlier study
(Kochanska & Murray, 2000). Children were shown
two sets of stories with vignettes: four by Eisenberg
(Eisenberg-Berg & Hand, 1979) and four by NunnerWinkler (1993), depicting protagonists matching the
child’s gender.
All Eisenberg story stems presented conflicts
between the protagonist’s interests and those of
others (e.g., ‘Birthday’ – run to a party or help an
injured child). E asked the child what the protagonist
should do and why, challenged the decision (pointed
out the needs of the other if the child chose the
selfish option, or of the protagonist if the child chose
the prosocial option), and asked the child again to
make a final choice. The Nunner-Winkler stories
depicted conflicts between the protagonist’s wishes
and those of others (e.g., ‘Drink’ – drink one’s soda or
share with a thirsty child).
Coding. We coded the child’s solutions to the
dilemmas. In the Eisenberg stories, each solution
was coded as selfish or prosocial. For each story, and
for both selfish and prosocial solutions, we used the
codes of 0, 1, 2, or 3. Zero was coded when the given
solution was absent in the child’s response; 1 when
the given solution was the child’s first choice but
changed when challenged, not final; 2 when it was
the child’s second or changed choice, retained as final; and 3 when it was the child’s first choice, unchanged when challenged, retained as final. In the
Nunner-Winkler stories, each solution was coded as
antisocial or as moral. Reliability, kappas, were .80
and .95 in the Eisenberg and Nunner-Winkler stories, respectively.
Data reduction. The scores were tallied across all
four stories in each set, to create two scores for the
Eisenberg stories: selfish and prosocial, and two
scores for the Nunner-Winkler stories: antisocial and
moral. The selfish score and the antisocial score
were summed into an overall selfish/antisocial
score, and the prosocial score and the moral score
into an overall prosocial/moral score. Finally, we
subtracted the former from the latter to create an
overall internalized moral judgment score at
56 months.
All descriptive data are in Table 1. For brevity, only
some of the composite variables were decomposed to
their constituent raw scores; all data are available
from the first author.
Mutually responsive orientation
25
Table 1 Descriptive data for the predictor, mediator, and outcome measures
Measure
M
SD
Range
a
Mutually responsive orientation, MRO
9 months
14 months
22 months
Overall MRO, 9–22 months
Child enjoyment of interaction, 33 months
Child committed compliance, 33 months
Maternal power assertionb, 33 months
Child conscience
Moral emotionc: guilt, 45 months
Moral conductd, 56 months
Moral cognitione, 56 months
.00
.00
.00
.01
1.16
.90
.00
.76
.72
.69
.57
.16
.16
.64
.00
.00
4.50
.77
.67
7.02
)1.80–1.76
)1.90–1.49
)1.88–1.85
)1.53–1.39
.68–1.68
.18–1.00
).35–3.50
)1.20–3.33
)3.00–.81
)14.00–16.00
a
Composite of standardized responsiveness and shared positivity.
Composite of standardized global power scores and physical power scores.
c
Composite of standardized scores: avoiding gaze, bodily tension, overall distress, negative affect, and positive affect (reversed).
d
Mean of internalized behavior while alone with prohibited toys and in the cheating game.
e
Difference between the prosocial/moral and selfish/antisocial cognition scores.
N’s: 9 months, 112; 14 months, 108; 22 months, 106; 33 months, 104; 56 months, 74.
b
Results
Early MRO and children’s future moral emotion,
conduct, and cognition: testing mediational
pathways
Preliminary analyses. Univariate correlations. The
correlations are in Table 2. All measures were
meaningfully correlated. MRO at 9–22 months predicted all three mediators at 33 months: child enjoyment of interaction with the mother, committed
compliance, and mother power assertion. MRO
also predicted all three measures of conscience:
45-month moral emotion, and 56-month conduct
and cognition.
The relations between the mediators and the
outcomes were mixed. The child’s enjoyment of
interaction with mother at 33 months predicted
future moral conduct and cognition, but not guilt.
Committed compliance at 33 months predicted
future guilt and moral conduct, but not cognition.
Maternal power at 33 months predicted the child’s
future guilt (marginally), moral conduct, but not
cognition. Further, the three mediators and the three
components of child conscience, respectively, were
significantly inter-correlated.
Gender differences. We tested gender differences in
the mediators and the outcomes. Whenever they
emerged, gender was controlled as a covariate in all
subsequent analyses. MANOVA for the three mediators at 33 months showed a marginal multivariate
effect of gender, F(3,100) ¼ 2.46, p < .07. Girls tended to enjoy interaction with the mothers more than
boys, F(1,102) ¼ 3.18, p < .08, girls M ¼ 1.19,
SD ¼ .16, boys M ¼ 1.14, SD ¼ .15, and girls were
higher than boys in committed compliance,
F(1,102) ¼ 6.18, p < .025, girls M ¼ .94, SD ¼ .11,
boys M ¼ .86, SD ¼ .19.
MANOVA for the three conscience measures indicated a near-significant multivariate effect of gender,
F(3,70) ¼ 2.70, p ¼ .05. Girls had higher scores than
boys on guilt, F(1,72) ¼ 6.28, p < .025, girls
M ¼ .31, SD ¼ .93, boys M ¼ ).17, SD ¼ .65, and on
moral conduct, F(1,72) ¼ 4.52, p < .05, girls
M ¼ .15, SD ¼ .53, boys M ¼ ).18, SD ¼ .78.
We also examined gender differences in MRO.
There were no differences in MRO between the dyads
with girls and those with boys, at any age, all t’s < 1.
The testing of mediation. These analyses tested
three conditions for mediation using a series of
Table 2 Correlations among the predictor, mediator, and outcome measures
Predictor (9–22 Mo)
Mediators (33 Mo)
Outcome measures
Mother–child
Child
Child
Mother
MRO
Enjoyment of
interaction
Committed
compliance
Power
assertion
Moral
emotion 45 mo.
Moral conduct
56 mo.
Moral
cognition 56 mo.
.26***
.26***
.30***
–
).35****
).35****
).59****
–
.30***
.11
.25**
).18+
–
.24*
.39****
.54****
).44****
.36****
–
.27**
.29**
.10
).09
.29**
.35***
MRO
Child enjoyment of interaction
Child committed compliance
Mother power assertion
Child moral emotion
Child moral conduct
*p < .05; **p < .025; ***p < .01; ****p < .001.
Child conscience
26
Grazyna Kochanska et al.
Table 3 The testing of the three conditions for mediation: standardized regression coefficients (Betas) from the separate regressions
Predictor (9–22 Mo)
Mother–child
MRO
Mediators (33 mo)
Dependent variables
Child
Child
Mother
Enjoyment of
interaction
Committed
compliance
Power
assertion
Child conscience
Moral
emotion 45 mo.
Moral
conduct 56 mo.
Moral
cognition 56 mo.
First condition: MRO at 9–22 months predicting the mediatorsa
MRO
.20*
.22**
).31***
–
–
–
Second condition: MRO at 9–22 months predicting child conscienceb
MRO
–
–
–
.20*
.22*
.27**
Third condition: the mediators at 33 months predicting child conscience, with a simultaneous decrease in the effect of MRO at
9–22 monthsc
Child enjoyment/interaction
–
–
–
).01
.33***
.25*
MRO
–
–
–
.22*
.15
.21+
Child committed compliance
MRO
–
–
–
–
–
–
.11
.18+
Mother power assertion
MRO
–
–
–
–
–
–
.01
.18+
.46****
.07
).04
.23+
).36***
.02
.06
.21
*p < .05; **p < .025; ***p < .01; ****p < .001.
a
The reported Betas come from the separate three regressions, each predicting one mediator at 33 months. The regressions
controlled for the earlier measures of the mediator (the composites of the 14- and 22-month scores), and gender when necessary
(predicting child enjoyment of interactions and committed compliance).
b
The reported Betas come from the separate three regressions, each predicting one conscience measure. The regressions controlled
for the earlier measures of conscience when available (the 22-month score for moral emotion, the composite of 14- and 22-month
scores of moral conduct), and gender when necessary (predicting moral emotion and moral conduct).
c
The reported Betas come from the separate three regressions for each conscience measure. Each of the mediators was entered along
with MRO. The regressions controlled for the earlier measures of conscience when available (the 22-month score for moral emotion,
the composite of 14- and 22-month scores of moral conduct), for the earlier measure of the mediator (the composite of the 14- and
22-month scores), and gender when necessary (predicting child moral emotion and conduct).
regressions (Baron & Kenny 1986).2 To simplify the
presentation, Table 3 includes only the Beta values
for the essential effects, from the final equations, and
the details are reported in the text.
Before testing the separate relations, we conducted an omnibus multivariate multiple regression to
assure that there was a significant overall relation
between the set of the three mediators and the set of
three outcome measures (SPSS-X User’s Guide,
1988, pp. 623–625). Those two sets of measures
were significantly related, F(9,210) ¼ 4.14, p < .001.
The first condition. We performed three multiple
regressions to test if the independent variable, MRO,
at 9–22 months, predicted each of the 33-month
mediators (child enjoyment of interaction with
mother, committed compliance, and maternal
power assertion). Only the effect of MRO on each
mediator is listed in Table 3 (reported as the first
row of Betas).
Because there was at least a marginal gender
effect for the first two mediators, in those two
2
Further to reduce the possibility that the mediational findings
may be due to an overlap between MRO (particularly the
shared positivity component) and children’s positive mood, we
conducted all the analyses, having removed the shared positivity component from the MRO score. The findings remained
substantially unchanged and thus, these analyses confirmed
that the findings cannot be attributed to any potential overlap.
Table 4 Comprehensive prediction of conscience outcomes
Predictors
t
Beta
Dependent variable: moral emotion at 45 months
Predictors entered in Step 1
Child gender
2.49**
).23
2.69***
.25
Earlier measure of child moral emotiona
Significant predictors retained in the equation
MRO, 9–22 months
2.16*
.20
Overall R2 ¼ .21, F (3,97) ¼ 8.66****
Dependent variable: moral conduct at 56 months
Predictors entered, Step 1
Child gender
1.19
).12
Earlier measure of child moral conductb
<1
.04
Significant predictors retained in the equation
Committed compliance, 33 months
4.08****
.43
Child enjoyment of interaction, 33 months 2.03*
.21
Overall R2 ¼ .35, F (4,69) ¼ 9.30****
Dependent variable: moral cognition at 56 monthsc
Significant predictors retained in the equation
Child enjoyment of interaction, 33 months 2.57**
.29
Overall R2 ¼ .08, F (1,72) ¼ 6.57**
Note: Beta and t values represent the final values, with all the
predictors in the equation. The predictors at Step 1 (child
gender, if applicable, and the earlier measure of the conscience
outcome, if available) were forced into the equation; the
remaining predictors (all three mediators, their earlier measures, and MRO) were evaluated stepwise. The predictors that
were retained are listed.
a
The 22-month score. bAn aggregate of the 14- and 22-month
scores. cFor moral cognition, there was no gender effect, and
no measure at 14–22 months was available. Thus, all the
predictors were evaluated stepwise.
+p < .10; *p < .05; **p < .025; ***p < .01; ****p < .001.
Mutually responsive orientation
regressions we also entered gender in the first step.
Further, to assure that the effect of MRO on each
mediator reflected a ‘real’ longitudinal change in the
mediator, and not simply continuity of concurrently
associated measures, in the first step, in each
regression, we controlled for the earlier measure of
the given mediator. Those earlier measures were,
respectively, the composites of the 14-and 22-month
scores of child enjoyment of interactions with the
mother, committed compliance, and the mother’s
power assertion. Indeed, although not reported in
Table 3, those earlier measures of the mediators
made significant contributions in the first two
regressions.
The first condition for mediation was met for all
three mediators. The 9–22-month MRO was a significant (positive) predictor of each of the 33-month
mediators. Children with histories of higher MRO
during the first two years enjoyed interacting with
their mothers more and showed more committed
compliance with them (even above the initially high
levels of those mediators), and the mothers resorted
to less power assertion at 33 months.
The second condition. We conducted three regressions (one for each component of conscience), to test
if the independent variable, 9–22-month MRO, predicted each dependent variable (moral emotion at
45 months, moral conduct and moral cognition at
56 months). Again, only those effects are in Table 3
(reported as the second row of Betas).
Because there was a significant gender effect for the
first two components of conscience, gender was
entered in the first step in those equations. Again, to
assure that any effect of MRO reflected a longitudinal
change in conscience and not just continuity of
concurrently associated measures, we entered an
earlier measure of the dependent variable when
available (a 22-month score on moral emotion, a
composite of the 14- and 22-month scores on moral
conduct; moral cognition was not assessed in the first
two years). Though not reported, the 22-month moral
emotion significantly predicted the 45-month score.
The second condition was met for all three conscience measures. Children with histories of higher
early MRO showed a stronger moral emotion of guilt
at 45 months, and moral conduct and moral cognition at 56 months. Each effect represented an increase over the earlier measure of the given outcome.
The third condition. Finally, for each of the three
conscience outcomes, we performed three regressions, one for each mediator. Each tested whether (a)
the given mediator was a significant predictor of the
conscience measure when considered together with
MRO, the independent variable, and (b) whether the
effect of MRO became not significant. Then, we
judged that mediation occurred. In the first step, we
controlled for child gender when necessary, and for
the earlier measures of both the given mediator and
27
the given conscience measure. The final effects
(only for a given mediator and MRO, for clarity) are
also in Table 3 (reported in the columns under each
outcome).
For children’s moral emotion, the third condition
for mediation was not met for any of the mediators.
Child gender and the earlier measures of guilt both
explained significant variance in children’s guilt at
45 months (not reported in Table 3). None of the
three mediators was a significant predictor. MRO
continued to be significant in the first equation, and
marginally significant in the other two, suggesting
that its effect was direct and not mediated.
For children’s moral conduct, the third condition
was met for all three mediators. Children’s enjoyment of interaction with mothers at 33 months,
committed compliance, and mothers’ power assertion all continued to predict children’s moral conduct at 56 months. Moreover, in all three equations,
the effect of MRO became non-significant.
For children’s moral cognition, the third condition
was met for only one mediator, children’s enjoyment
of interaction with the mothers at 33 months, which
continued to predict moral cognition, with the effect
of MRO dropping to the marginal level. Committed
compliance and maternal power assertion were not
significant predictors of moral cognition.
The testing of all three mediators. Finally, for each
conscience outcome measure, we conducted a
comprehensive multiple regression where all three
mediators were examined simultaneously. In the first
step, gender (if necessary) and the earlier measure of
the conscience outcome (if available) were entered. To
allow for a clear view of the significant effects, and
given that all the effects had been examined
separately in the previous regressions, here, after the
first step, all the remaining predictors (MRO, all three
mediators, and all their earlier measures) were
allowed to compete, and were evaluated in the
stepwise manner. The findings are in Table 4.
For the moral emotion of guilt, the findings were
consistent with the earlier separate regressions.
After controlling for gender and guilt at 22 months,
both significant predictors, early MRO continued
directly to predict children’s guilt at 56 months.
There were no mediating effects.
For moral conduct, there were two significant
mediation effects consistent with the earlier regressions. The effect of MRO dropped to non-significant
level, but two mediators at 33 months – child committed compliance and enjoyment of interaction with
mother – remained significant. In contrast to the
earlier regressions, however, power assertion, when
tested together with the other two mediators, no
longer independently predicted moral conduct.
The results for moral cognition were consistent
with the separate regressions. Children’s enjoyment
of interaction with mothers mediated the relation
between MRO and this conscience outcome.
28
Grazyna Kochanska et al.
Discussion and conclusions
MRO and future conscience
Children’s conscience is rarely studied in developmental psychopathology, although conduct problems in children and adults can be seen as reflecting
impaired moral emotion, conduct, and cognition
(Frick et al., 2003; Lykken, 1995). We revisit the
classic question: Is early conscience development
unitary? Most often, this question has been asked in
structural terms – is conscience a global, coherent,
trait-like structure, such as superego or a ‘moral
stage’, or a set of unrelated or loosely related components (Burton, 1963; Hartshorne & May, 1928–
1930; Rushton et al., 1983; Sears et al., 1965)? Our
study informs this question by showing that moral
emotion, conduct, and cognition are significantly, if
modestly, coherent at preschool age (Table 2). The
major contribution of this work, however, goes beyond the structure of conscience. We ask if conscience development is a unitary process, or if its
components unfold along multiple pathways. Our
findings support the latter: Distinct paths appear to
link the early parent–child relationship with child
future moral emotion, conduct, and cognition.
This study affords new – albeit preliminary –
insights into the potential mediating mechanisms
responsible for the established empirical associations between mother–child mutually responsive
relationship and the child’s conscience. We again
replicated, in a new sample and at a new age, several
links: between mother–child MRO and committed
compliance, between MRO and maternal power
assertion, and between committed compliance and
conscience (Kochanska, 1997; Kochanska & Murray, 2000). We also demonstrated new links: the
longitudinal association between early MRO and the
child’s enjoyment of interactions with the mother,
and between that enjoyment and the child’s future
moral conduct and cognition. Those findings complement the research on parent–child synchrony,
positive reciprocity, and mutual positivity as factors
that decrease children’s future risk for conduct
problems (Criss, Shaw, & Ingoldsby, 2003; DeaterDeckard & Petrill, 2004; Gardner, Ward, Burton, &
Wilson, 2003; Shaw et al., 1994). The core new
contribution of this study, however, consistent with
the current zeitgeist (Eisenberg, 1998), was to take
the first step toward examining mechanisms that
link the mother–child MRO and child developing
conscience.
Children’s enjoyment of interactions with mothers
as a mediator of conscience. As expected, MRO
between the mother and the child, established during the first two years of life, enhanced children’s
enjoyment of their interactions at 33 months, controlling for the initial level of pleasure in interactions.
This enjoyment, in turn, fostered children’s moral
conduct and cognition at 56 months. Those findings
add to the large body of literature that has linked
temporary or dispositional positive emotions with
prosocial behavior and self-regulation (Carlson et al.,
1988; Cialdini, Kenrick, & Baumann, 1982; Eisenberg & Fabes, 1998; Fry, 1975; Isen, 1999; Lay et al.,
1989; Moore & Isen, 1990). Because we assessed
children’s enjoyment two years earlier than their
consciences, clearly, early mother–child relationship
fostered in the child a form of positive disposition
that was more enduring than a mere positive mood
during a sample of interactions with the mother.
More likely, it promoted the child’s deeper and lasting positive disposition toward the mother, which in
turn, over time, fostered his or her eagerness to
embrace the mother’s rules and values.
Those findings echo Maccoby’s (1983) emphasis
on the role of positive affect in socialization. She
proposed that research on links between positive
mood and prosocial behavior has important implications for socialization. She suggested that the
history of sensitive, playful, and cooperative interactions and the resulting positive affective bond, ‘…
engenders attention to the parent and a readiness to
become interested in, and cooperative with, an
enterprise suggested by the parent’ (Maccoby, 1983,
p. 363). Our findings also support and expand Lay
et al.’s (1989) conclusions: ‘The fact that maternal
responsiveness induces a positive mood points to a
mechanism that may be significant in both attachment and socialization beyond infancy’ (p. 1405).
From the attachment perspective, the child’s positive
emotions toward the parent may promote moral
development by fostering security, and by increasing
receptiveness to the parent and eagerness to adopt
her agenda. Children who are happy in their relationships with their parents may pay more attention
to them, look forward to interacting with them, feel
secure in the relationship, and thus be more eager
learners during the socialization process (Lay et al.,
1989; Maccoby, 1983; Waters et al., 1990).
To show that the child’s enjoyment of interactions
with the parent mediates the relations between MRO
and moral conduct and, more tentatively, moral
cognition, is only the first step. A challenge for future
research is to understand the mechanism or mechanisms of that relation.
Social psychology has proposed several mechanisms that may link positive emotions with moral
behavior and cognition and that may help explain
why children who enjoy their interactions with their
parents develop strong consciences, act in accord
with rules, and weigh prosocial values over selfish
concerns in their moral judgments. They may act to
maintain good feelings by engaging in ‘good’ behavior, itself a source of positive emotions (Waters et al.,
1990; Emde, 1991), consistent with the ‘positive
mood maintenance’ hypothesis (Carlson et al.,
1988). That hypothesis implies that the child who is
experiencing positive mood would seek to maintain
that emotional state. One way to do so is to engage in
Mutually responsive orientation
behavior consistent with parental rules, because
such ‘good’ behavior elicits positive emotion (Emde,
1991).
Those children may also have positive, benevolent,
and cooperative views of others and the society, and
thus endorse prosocial, non-selfish choices in their
moral judgments, consistent with the ‘social outlook’
hypothesis (Carlson et al., 1988). That hypothesis
assumes that good mood, particularly when it is due
to social stimuli and events, primes communal and
cooperative cognitions and perceptions. Such cognitions may promote children’s more generous, empathic, and prosocial thoughts and reasoning.
Happy mood may also reduce a child’s defensive focus on self (Moore & Isen, 1990), and thus enhance
prosocial conduct and the ability to consider others’
perspectives and welfare. They may feel less worried
about their own well-being and less defensive, and
feel more generous toward others (Moore & Isen,
1990). And finally, positive mood also increases the
capacity for restraint when dealing with temptations
(Eisenberg & Fabes, 1992; Fry, 1975; Seeman &
Schwarz, 1974). Without concluding which specific
mechanism links the child’s enjoyment of interactions with the parent with the developing conscience,
we have shown such a link exists, and, further, that
such enjoyment mediates the links between early
MRO and future moral conduct and cognition.
Because the moral emotion of guilt is associated
with negative rather than positive affect, we had not
anticipated – and indeed did not find – children’s
enjoyment of interactions with mothers to mediate
the link between MRO and guilt. Children who had
been in highly mutually responsive early relationships, as preschoolers were more prone to experience guilt following transgressions, but this effect of
MRO was direct. It is another challenge for future
research to determine what mechanisms may be
responsible for that impact. Perhaps other mediators, not studied here, link the early mother–child
relationship with child guilt (Zahn-Waxler & Kochanska, 1990). For example, MRO may foster the
child’s high self-esteem, and consequently, increase
guilt and remorse when his or her behavior falls
short of standards.
Future research will need to reconcile the findings
on positive emotion and moral conduct and cognition
with the research on negative emotions, such as fear
or sadness, and aspects of conscience, such as guilt
or empathy (Kagan, 1998; Kochanska et al., 1994;
Rothbart, Ahadi, & Hershey, 1994; Rothbart &
Bates, 1998). Ultimately, an integration of those literatures will produce a comprehensive theory of
positive and negative emotions, both transient and
enduring, as factors that influence moral emotions,
conduct, and cognition.
Children’s committed compliance as a mediator of
conscience. As expected, high MRO in the mother–
child dyad during the first two years promoted future
29
conscience through another mediational mechanism: through the child’s committed compliance to
the mother. This mediation, however, was only true
for children’s moral conduct. MRO fostered children’s wholehearted, eager, self-regulated compliance with maternal prohibitions at 33 months, and
it, in turn, promoted moral conduct at 56 months,
but had no effect on moral emotion or cognition. Our
findings support others’ view of willing compliance
as an antecedent of internalization (Emde et al.,
1991; Londerville & Main, 1981; Maccoby & Martin,
1983), and we elucidate the developmental mediational chain. Several mechanisms accounting for
that mediation are possible. Maternal responsiveness and shared positive affect between the mother
and child promote the child’s cooperative attitude,
characterized by an internal sense of obligation
(Maccoby, 1999), and thus, they compel the young
child to engage in committed compliance. Because of
the voluntary nature of committed compliance,
children likely attribute their behavior to their own
wishes, and experience a sense of choice and
autonomy, which in turn leads to the integration of
the particular behavioral rule with their own selves,
or genuine internalization (Grolnick, Deci, & Ryan,
1997; Grusec & Goodnow, 1994; Lepper, 1981).
Over time, children who engage in committed compliance may come to view themselves as ‘good’ and
moral, and such ‘moral self’ increasingly serves as a
guide for behavior (Kochanska, 2002b), consistent
with attributional and self-determination theories
(Ryan & Deci, 2000). Again, regardless of the specific
mechanism involved, we expected that committed
compliance would serve as a mediator linking MRO
with children’s moral emotion, moral conduct, and
moral cognition.
Why did committed compliance mediate only the
link between MRO and moral conduct? Perhaps it
was because the nature of the rules with which the
children had complied at 33 months was quite similar to that involved in the internalization paradigms
at 56 months. All those rules were proscriptive
‘Don’ts’, specifying what behavior was forbidden,
and requiring restraint.
Future research should address a possibility that,
with age, the mediational effects of committed compliance may gradually generalize beyond such ‘isomorphic’ behavior. If indeed the experience of
committed compliance leads children to view themselves as moral and ‘good’ (Kochanska, 2002b), then
as they grow older, those self-attributions may become more general. Children may begin to apply
them to a broader moral spectrum, including moral
emotion and cognition. Grusec and Redler (1980)
found that attributions of helpfulness to the self increased generalized altruism, but not before age 8.
Although the families in this study represented a
low-risk child-rearing environment, our findings are
relevant to developmental psychopathology and
they may inform our understanding of pathways
30
Grazyna Kochanska et al.
undermining conscience development in children
in at-risk environments. For example, physically
abused children have shown less committed compliance than children in non-maltreating families
(Koenig, Cicchetti, & Rogosch, 2000).
Maternal power assertion as a mediator of conscience. The marginal univariate relation between
maternal power at 33 months and children’s future
emotion of guilt disappeared when power and MRO
were entered together (whether or not the other two
mediators were also entered). Maternal power
assertion was also unrelated to children’s future
moral cognition, a finding consistent with our recent
work on domain specificity regarding the effects of
power (Kochanska, Aksan, & Nichols, 2003).
Power assertion, when tested alone, did mediate
between MRO and children’s moral conduct at
56 months. However, when tested together with
MRO and the other two mediators – child enjoyment
of interaction with mother and committed compliance – maternal power was no longer a mediator of
that link. These intriguing findings add to the growing discussion of parental power assertion in the
context of early relationships. Several non-mutually
exclusive interpretations are possible. Perhaps the
most important is the fact that the mediators shared
common variance. Children who enjoyed their
interactions with mothers also showed more committed compliance, and received less maternal
power. It is possible – although entirely beyond the
scope of this article – that the three mediators were
causally related over time. As just one example of a
possible direction of effect, children who had enjoyed
their interactions with their mothers may become
more eager to comply, thus causing their mothers to
use less power assertion. In our final regression, we
did not examine the causal links among the mediators. In future research, with larger samples, structural equations techniques could help model the
potentially more complex causal effects, and thus
allow for a more appropriate examination of the
developmental processes involved.3
Another possibility involves heterogeneity within
the relatively amorphous category of power assertion
(Grusec & Goodnow, 1994). Mothers who have
formed high MRO with their children and whose
children happily cooperate, when using power at all,
may resort to firm, clear commands and guidance
that may be easily accepted by the child and promote
conscience. Mothers who have failed to form such
3
Several factors argued against presenting the results using
structural equation modeling. The most notable reason was
the ratio of sample size to the number of estimated parameters
in the full model. A sample of 74 was highly inadequate in
obtaining stable maximum likelihood estimates. Furthermore,
the tests of the smaller components of the comprehensive
model (for each conscience outcome separately) did not suggest different conclusions from those presented in the article.
responsive relationships, and whose children are
more difficult, may use more confrontational, sharp,
or angry strategies, which may be resented by the
child and undermine conscience. A more finegrained coding of power assertion might provide a
window into such a possibility.
A third possibility involves the issue of the context
in which power is used. Its maladaptive effects have
not been found across all ecologies of parenting.
Patterson (Patterson, DeBaryshe, & Ramsey, 1989)
found repeatedly that in family relationships that are
generally adversarial, children do not respond to
punishment, in contrast to children in more positive
relationships. Baumrind (1983) proposed that firm
discipline applied in the context of an accepting and
warm relationship promotes adaptive outcomes.
Darling and Steinberg (1993) argued that parenting
style serves as a relationship context which moderates the effectiveness of discipline. In some cultures
or sub-cultures, power assertion is viewed as a
component of a loving, child-oriented relationship,
and in those populations, it does not have negative
implications for socialization (Chao, 1994; DeaterDeckard, Dodge, Bates, & Pettit, 1996). Finally, in
our study, the range of maternal power was relatively
constrained; most mothers resorted to powerassertive strategies rarely, and when they did, they
used mostly low-intensity tactics. In contrast, the
other two mediators were well distributed. Our failure to support the mediational effect of power may
have been due to this problem. Future exploration of
these issues in populations where the range of parental discipline techniques is broader, for example in
child-maltreating families, seems warranted.
These considerations support a need for more research on power assertion (Grusec & Goodnow,
1994). It would be currently premature to conclude
that power per se does not interfere with the early
development of conscience. Its effects may be indirect, and may depend on the context, population,
and other characteristics of the child and the relationship with the parent.
A useful contribution of this study is another
demonstration of the convergence between microand macroscopic responsiveness, and between responsiveness and shared positivity. Deater-Deckard
and O’Connor (2000) replicated the structure of MRO
in two samples, relying on ratings rather than
microscopic coding. We have also again demonstrated longitudinal continuity of MRO, beyond continuity from toddler to preschool age in another
sample (Kochanska, 1997), and from 9 to 14 months
in this sample (Kochanska et al., 1999).
Limitations and future directions
This study has several limitations. Most importantly,
we need to exercise caution regarding causal inferences, even when our longitudinal analyses, which
controlled for the earlier levels of the constructs,
Mutually responsive orientation
supported the existence of mediational paths. In
non-experimental designs, competing interpretations are always possible. For example, recent
advances in behavioral genetics suggest some genetic effects in parent–child mutuality (Deater-Deckard & O’Connor (2000), along with nonshared effects
(Deater-Deckard et al., 2001).
In view of the increasingly recognized importance
of the network of family relationships (Lamb, 1997;
Parke & Buriel, 1998), our exclusive focus on
mothers is another limitation. The role of fathers in
the development of children’s conscience remains
unexplored, and is one of the urgent challenges for
future research.
How early conscience is formed, whether its
development progresses along a unitary path or
multiple paths, whether and how mother–child
relationship influences its development, and
whether early measures of conscience predict future
behavior problems are compelling questions of
development and psychopathology. This research
adds to our search for the answers.
Acknowledgements
This research was supported by the NSF grant, SBR9510863, the NIMH Independent Scientist Award,
KO2 MH01446, and RO1 MH63096 to the first author. We thank Pui-Wa Lei for her assistance with
the analyses, many students and staff for conducting the sessions and coding the data, and the families in the Parent–child Study for their outstanding
commitment to this research.
Correspondence to
Grazyna Kochanska, Department of Psychology,
University of Iowa, Iowa City, IA 52242-1407, USA;
Tel: (319) 335 2409; Fax: (319) 335 0191; Email:
[email protected]
References
Ainsworth, M.D.S., Bell, S.M., & Stayton, D.J. (1971).
Individual differences in Strange Situation behaviour
of one-year-olds. In H.R. Schaffer (Ed.), The origins of
human social relations (pp. 17–57). London: Academic Press.
Baron, R.M., & Kenny, D.A. (1986). The moderator–
mediator variable distinction in social psychological
research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.
Baumrind, D. (1983). Rejoinder to Lewis’ reinterpretation of parental firm control effects: Are authoritative
families really harmonious? Psychological Bulletin,
94, 132–142.
Belsky, J. (1999). Interactional and contextual determinants of attachment security. In J. Cassidy & P.R.
31
Shaver (Eds.), Handbook of attachment: Theory,
research, and clinical applications. New York: Guilford Press.
Blair, R.J.R. (1997). Moral reasoning and the child with
psychopathic tendencies. Personality and Individual
Differences, 22, 731–739.
Bretherton, I., Golby, B., & Cho, E. (1997). Attachment
and the transmission of values. In J.E. Grusec & L.
Kuczynski (Eds.), Parenting and children’s internalization of values: A handbook of contemporary theory
(pp. 103–134). New York: Wiley.
Bryant, B.K., & Crockenberg, S.B. (1980). Correlates
and dimensions of prosocial behavior: A study of
female siblings with their mothers. Child Development, 51, 529–544.
Bugental, D.B. (2000). Acquisition of the algorithms of
social life: A domain-based approach. Psychological
Bulletin, 126, 187–219.
Bugental, D.B., & Goodnow, J.J. (1998). Socialization
processes. In W. Damon (Series Ed.), Handbook of
child psychology, N. Eisenberg (Vol. Ed.), Social,
emotional, and personality development (5th edn,
pp. 389–462). New York: Wiley.
Burton, R.V. (1963). The generality of honesty reconsidered. Psychological Review, 70, 481–499.
Carlson, M., Charlin, V., & Miller, N. (1988). Positive
mood and helping behavior: A test of six hypotheses.
Journal of Personality and Social Psychology, 55,
211–229.
Chao, R.K. (1994). Beyond parental control and authoritarian parenting style: Understanding Chinese
parenting through the cultural notion of training.
Child Development, 65, 1111–1119.
Cialdini, R.B., Kenrick, D.T., & Baumann, D.J. (1982).
Effects of mood on prosocial behavior in children and
adults. In N. Eisenberg (Ed.), The development of
prosocial behavior (pp. 339–359). New York: Academic Press.
Clark, M.S. (1984). Record keeping in two types of
relationships. Journal of Personality and Social Psychology, 47, 549–557.
Collins, W.A., & Laursen, B. (Eds.) (1999). Relationships as developmental contexts. Minnesota Symposia
on Child Psychology, 30, Hillsdale, NJ: Erlbaum.
Collins, W.A., Maccoby, E.E., Steinberg, L., Hetherington, E.M., & Bornstein, M.H. (2000). Contemporary
research on parenting: The case for nature and
nurture. American Psychologist, 55, 218–232.
Criss, M.M., Shaw, D.S., & Ingoldsby, E.M. (2003).
Mother–son positive synchrony in middle childhood:
Relation to antisocial behavior. Social Development,
12, 379–400.
Darling, N., & Steinberg, L. (1993). Parenting style as
context: An integrative model. Psychological Bulletin,
113, 487–496.
Deater-Deckard, K., & O’Connor, T.G. (2000). Parent–
child mutuality in early childhood: Two behavioral
genetic studies. Developmental Psychology, 36, 561–
570.
Deater-Deckard, K., Dodge, K.A., Bates, J.E., & Pettit,
G.S. (1996). Physical discipline among African American and European American mothers: Links to
children’s externalizing behaviors. Developmental
Psychology, 32, 1065–1072.
32
Grazyna Kochanska et al.
Deater-Deckard, K., & Petrill, S.A. (2004). Parent–child
dyadic mutuality and child behavior problems:An
investigation of gene–environment process. Journal of
Child Psychology and Psychiatry, 45, 1171–1179.
Deater-Deckard, K., Pike, A., Petrill, S.A., Cutting, A.L.,
Hughes, C., & O’Connor, T. (2001). Nonshared environmental processes in social-emotional development:
An observational study of identical twin differences in
the preschool period. Developmental Science, 4,
F1–F6.
De Wolff, M.S., & van IJzendoorn, M.H. (1997).
Sensitivity and attachment: A meta-analysis of parental antecedents of infant attachment. Child Development, 68, 571–591.
Dix, T. (1991). The affective organization of parenting:
Adaptive and maladaptive processes. Psychological
Bulletin, 110, 3–25.
Eisenberg, N. (1998). Introduction. In W. Damon (Series
Ed.), Handbook of child psychology, N. Eisenberg
(Vol. Ed.), Social, emotional, and personality development (5th edn, pp. 1–24). New York: Wiley.
Eisenberg, N., & Fabes, R.A. (1998). Prosocial development. In W. Damon (Series Ed.), Handbook of child
psychology, N. Eisenberg (Vol. Ed.), Social, emotional,
and personality development (pp. 701–778). New
York: Wiley.
Eisenberg-Berg, N., & Hand, M. (1979). The relationship of preschoolers’ reasoning about prosocial moral
conflicts to prosocial behavior. Child Development,
50, 356–363.
Emde, R.N. (1991). Positive emotions for psychoanalytic
theory – surprises from infancy research and new
directions. Journal of the American Psychoanalytic
Association, 39, 5–44 Suppl.
Emde, R.N., Biringen, Z., Clyman, R.B., & Oppenheim,
D. (1991). The moral self of infancy: Affective core and
procedural knowledge. Developmental Review, 11,
251–270.
Frick, P.J., Cornell, A.H., Bodin, S.D., Dane, H.E.,
Barry, C.T., & Loney, B.R. (2003). Callous-unemotional traits and developmental pathways to severe
conduct problems. Developmental Psychology, 39,
246–260.
Fry, P.S. (1975). Affect and resistance to temptation.
Developmental Psychology, 11, 466–472.
Gardner, F., Ward, S., Burton, J., & Wilson, C. (2003).
The role of mother–child joint play in the early
development of children’s conduct problems: A longitudinal study. Social Development, 12, 361–378.
Grolnick, W.S., Deci, E.L., & Ryan, R.M. (1997).
Internalization within the family: The self-determination theory perspective. In J.E. Grusec & L. Kuczynski (Eds.), Parenting and children’s internalization of
values. A handbook of contemporary theory (pp. 135–
161). New York: Wiley.
Grusec, J.E. (1997). A history of research on parenting strategies and children’s internalization of values. In J.E. Grusec & L. Kuczynski (Eds.),
Parenting and children’s internalization of values.
A handbook of contemporary theory (pp. 3–22). New
York: Wiley.
Grusec, J.E., & Goodnow, J.J. (1994). The impact of
parental discipline methods on the child’s internalization of values: A reconceptualization of the current
points of view. Developmental Psychology, 30, 4–19.
Grusec, J.E., & Redler, E. (1980). Attribution, reinforcement, and altruism: A developmental analysis.
Developmental Psychology, 16, 525–534.
Harrist, A.W., & Waugh, R.M. (2002). Dyadic synchrony: Its structure and function in children’s
development. Developmental Review, 22, 555–592.
Hartshorne, H., & May, M.S. (1928–1930). Studies in
the nature of character. New York: Macmillan.
Hartup, W.W. (1989). Social relationships and their
developmental significance. American Psychologist,
44, 120–126.
Hinde, R.A. (1979). Towards understanding relationships. New York: Academic Press.
Hoffman, M.L. (1983). Affective and cognitive processes
in moral internalization. In E.T. Higgins, D. Ruble, &
W.W. Hartup (Eds.), Social cognition and social
development (pp. 236–274). New York: Cambridge
University Press.
Isen, A.M. (1999). Positive affect. In T. Dalgleish & M.J.
Power (Eds.), Handbook of cognition and emotion (pp.
521–539). Chichester, England: Wiley.
Kagan, J. (1998). Biology and the child. In W. Damon
(Series Ed.), Handbook of child psychology, N. Eisenberg (Vol. Ed.), Social, emotional, and personality
development (5th edn, pp. 177–236). New York: Wiley.
Kochanska, G. (1993). Toward a synthesis of parental
socialization and child temperament in early development of conscience. Child Development, 64, 325–
347.
Kochanska, G. (1997). Mutually responsive orientation
between mothers and their young children: Implications for early socialization. Child Development, 68,
94–112.
Kochanska, G. (2002a). Mutually responsive orientation between mothers and their young children: A
context for the early development of conscience.
Current Directions in Psychological Science, 11, 191–
195.
Kochanska, G. (2002b). Committed compliance, moral
self, and internalization: A mediational model. Developmental Psychology.
Kochanska, G., & Aksan, N. (1995). Mother–child
mutually positive affect, the quality of child compliance to requests and prohibitions, and maternal
control as correlates of early internalization. Child
Development, 66, 236–254.
Kochanska, G., Aksan, N., & Koenig, A.L. (1995). A
longitudinal study of the roots of preschoolers’ conscience: Committed compliance and emerging internalization. Child Development, 66, 1752–1769.
Kochanska, G., Aksan, N., & Nichols, K.E. (2003).
Maternal power assertion in discipline and moral
discourse contexts: Commonalities, differences, and
implications for children’s moral conduct and cognition. Developmental Psychology, 39, 949–963.
Kochanska, G., Coy, K.C., & Murray, K.T. (2001). The
development of self-regulation in the first four years
of life. Child Development, 72, 1091–1111.
Kochanska, G., Coy, K.C., Tjebkes, T.L., & Husarek,
S.J. (1998a). Individual differences in emotionality in
infancy. Child Development, 69, 375–390.
Kochanska, G., DeVet, K., Goldman, M., Murray, K.T.,
& Putnam, S.P. (1994). Maternal reports of conscience development and temperament in young
children. Child Development, 65, 852–868.
Mutually responsive orientation
Kochanska, G., Forman, G., & Coy, K.C. (1999).
Implications of the mother–child relationship in
infancy for socialization in the second year of life.
Infant Behavior and Development, 22, 249–265.
Kochanska, G., Gross, J.N., Lin, M.-H., & Nichols, K.E.
(2002). Guilt in young children: Development, determinants, and relations with a broader system of
standards. Child Development, 73, 461–482.
Kochanska, G., & Murray, K.T. (2000). Mother–child
mutually responsive orientation and conscience
development: From toddler to early school age. Child
Development, 71, 417–431.
Kochanska, G., Padavich, D.L., & Koenig, A.L. (1996).
Children’s narratives about hypothetical moral dilemmas and objective measures of their conscience:
Mutual relations and socialization antecedents. Child
Development, 67, 1420–1436.
Kochanska, G., Tjebkes, T.L., & Forman, D.R.
(1998b). Children’s emerging regulation of conduct:
Restraint, compliance, and internalization from
infancy to the second year. Child Development, 69,
1378–1389.
Koenig, A.L., Cicchetti, D., & Rogosch, F.A. (2000).
Child compliance/noncompliance and maternal contributors to internalization in maltreating and nonmaltreating dyads. Child Development, 71, 1018–
1032.
Laible, D.J., & Thompson, R.A. (2000). Mother–child
discourse, attachment security, shared positive affect, and early conscience development. Child Development, 71, 1424–1440.
Lamb, M.E. (Ed.) (1997). The role of the father in child
development. New York: Wiley.
Lay, K.L., Waters, E., & Park, K.A. (1989). Maternal
responsiveness and child compliance: The role of
mood as mediator. Child Development, 60, 1405–
1411.
Lepper, M.R. (1981). Intrinsic and extrinsic motivation
in children: Detrimental effects of superfluous social
controls. In W.A. Collins (Ed.), Minnesota symposia
on child psychology (vol. 14, pp. 155–214). Minneapolis: University of Minnesota Press.
Londerville, S., & Main, M. (1981). Security of attachment, compliance, and maternal training methods in
the second year of life. Developmental Psychology, 17,
289–299.
Lykken, D.T. (1995). The antisocial personalities. Hillsdale, NJ: Erlbaum.
Lytton, H. (1980). Parent–child interaction: The socialization processes observed in twin and singleton
families. New York: Plenum.
Maccoby, E.E. (1983). Let’s not overattribute to the
attribution process: Comments on social cognition
and behavior. In E.T. Higgins, D.N. Ruble, & W.W.
Hartup (Eds.), Social cognition and social development
(pp. 356–370). New York: Cambridge University
Press.
Maccoby, E.E. (1992). The role of parents in the
socialization of children: An historical overview.
Developmental Psychology, 28, 1006–1017.
Maccoby, E.E. (1999). The uniqueness of the parent–
child relationship. In W.A. Collins & B. Laursen
(Eds.), Relationships as developmental contexts. Minnesota symposia on child psychology 30 (pp. 157–
175). Hillsdale, NJ: Erlbaum.
33
Maccoby, E.E., & Martin, J.A. (1983). Socialization in
the context of the family: Parent–child interaction. In
P.H. Mussen (Series Ed.), E.M. Hetherington (Vol.
Ed.), Handbook of child psychology, Vol. 4, Socialization, personality, and social development (4th edn, pp.
1–102). New York: Wiley.
Martin, J.A. (1981). A longitudinal study of the consequences of early mother–infant interaction: A microanalytic approach. Monographs of the Society for
Research in Child Development, 46, (3, Serial No.
190).
Matas, L., Arend, R.A., & Sroufe, L.A. (1978). Continuity
of adaptation in the second year: The relationship
between quality of attachment and later competence.
Child Development, 49, 547–556.
Moore, B.S., & Isen, A. (1990). Affect and social
behavior. In B.S. Moore & A.M. Isen (Eds.), Affect
and social behavior (pp. 1–21). New York: Cambridge
University Press.
Nunner-Winkler, G. (1993). The growth of moral motivation. In G.G. Noam & T.E. Wren (Eds.), The moral
self (pp. 269–291). Cambridge, MA: MIT Press.
Patterson, G.R., DeBaryshe, B.D., & Ramsey, E. (1989).
A developmental perspective on antisocial behavior.
American Psychologist, 44, 329–335.
Parke, R.D., & Buriel, R. (1998). Socialization in the
family: Ethnic and ecological perspectives. In W.
Damon (Series Ed.), Handbook of child psychology,
N. Eisenberg (Vol. Ed.), Vol. 3., Social, emotional, and
personality development (5th edn, pp. 463–552). New
York: Wiley.
Parpal, M., & Maccoby, E.E. (1985). Maternal responsiveness and subsequent child compliance. Child
Development, 56, 1326–1334.
Radke-Yarrow, M., Richters, J., & Wilson, W.E. (1988).
Child development in a network of relationships. In
R.A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families: Mutual influences (pp. 48–67).
Oxford: Clarendon.
Radke-Yarrow, M., ZahnWaxler, C., & Chapman, M.
(1983). Children’s prosocial dispositions and behavior. In P. Mussen (Series Ed.) & E.M. Hetherington
(Ed.), Socialization, personality, and social development: Vol. 4. Handbook of child psychology (4th edn,
pp. 469–545). New York: Wiley.
Reis, H.T., Collins, W.A., & Berscheid, E. (2000).
Relationships in human behavior and development.
Psychological Bulletin, 126, 844–872.
Rocissano, L., Slade, A., & Lynch, V. (1987). Dyadic
synchrony and toddler compliance. Developmental
Psychology, 23, 698–704.
Rothbart, M.K., Ahadi, S.A., & Hershey, K.L. (1994).
Temperament and social behavior in childhood.
Merrill-Palmer Quarterly, 40, 21–39.
Rothbart, M.K., & Bates, J.E. (1998). Temperament. In
W. Damon (Series Ed.), Handbook of child psychology, N. Eisenberg (Vol. Ed.), Social, emotional, and
personality development (pp. 105–176). New York:
Wiley.
Rushton, J.P., Brainerd, C.J., & Pressley, M. (1983).
Behavioral development and construct validity: The
principle of aggregation. Psychological Bulletin, 94,
18–38.
Ryan, R.M., & Deci, E.L. (2000). Self-determination
theory and the facilitation of intrinsic motivation,
34
Grazyna Kochanska et al.
social development, and well-being. American Psychologist, 55, 68–78.
Sears, R.R., Rau, L., & Alpert, R. (1965). Identification
and child rearing. Stanford: Stanford University
Press.
Seeman, G., & Schwarz, J.C. (1974). Affective state and
preference for immediate versus delayed reward.
Journal of Research in Personality, 7, 384–394.
Shaw, D.S., Keenan, K., & Vondra, J.I. (1994). Developmental precursors of externalizing behavior: Ages 1
to 3. Developmental Psychology, 30, 355–364.
Stayton, D.J., Hogan, R., & Ainsworth, M.D.S. (1971).
Infant obedience and maternal behavior: The origins
of socialization reconsidered. Child Development, 42,
1057–1069.
Thompson, R.A. (1998). Early sociopersonality development. In W. Damon (Series Ed.), Handbook of child
psychology, N. Eisenberg (Vol. Ed.), Vol. 3, Social,
emotional, and personality development (pp. 25–104).
New York: Wiley.
Thompson, R.A. (2000). The legacy of early attachments. Child Development, 71, 145–152.
van Ijzendoorn, M.H. (1997). Attachment, emergent
morality, and aggression: Toward a developmental
socioemotional model of antisocial behaviour. International Journal of Behavioral Development, 21, 703–
727.
Waters, E., Kondo-Ikemura, K., Posada, G., & Richters,
J.E. (1990). Learning to love: Mechanisms and milestones. In M. Gunnar & L.A. Sroufe(Eds.), Minnesota
symposia on child psychology (Vol. 23, pp. 217–255).
Hillsdale, NJ: Lawrence Erlbaum.
Westerman, M.A. (1990). Coordination of maternal
directives with preschoolers’ behavior in compliance-problem and healthy dyads. Developmental
Psychology, 26, 621–630.
Zahn-Waxler, C., & Kochanska, G. (1990). The origins
of guilt. In R.A. Thompson (Ed.), The 36th annual
Nebraska symposium on motivation: Socioemotional
development (pp. 183–257). R. A. Dienstbier (Series
Ed.). Lincoln: University of Nebraska Press.
Manuscript accepted 5 February 2004