Infant emotion regulation with mothers and fathers

INFANT BEHAVIOR AND DEVELOPMENT 20 (l), 1997, pp. 47-57 ISSN 0163-6383

Copyright 0 1997 ABLEX Publishing Corporation All rights of reproduction in any form reserved.

Infant Emotion Regulation with Mothers and Fathers LISA J. BRIDGES

University of California, Riverside

WENDY S. GROLNICK

Clark Universig

JAMES P. CONNELL

Institute for Research and Reform in Education

This study was designed to examine infant emotion regulation within mildly frustrating delay situations with mothers and fathers. Contextual influences on, and consistent individual differences in, observed strategy use and expressed emotion were assessed in a sample of 12%.to 14.month- old infants, each seen on two occasions, once with each parent. Infants were seen in two delay situations with each parent, one in which the parent was asked to remain fairly inactive. and one in which no restrictions were placed on parental behavior. Results suggested strong contextual (parent-passive vs. parent-active) influences on the level of expressed negative versus positive emotion. and on the types of emotion regulation strategies used by infants, but few sex-of-parent effects. Overall, results suggest that expressed emotion and strategy use are more similar than different in interactions with mothers and fathers. However, some differential use of particular emotion regulation strategies may reflect developing adaptations made by infants to different characteristics of relationships with mothers and fathers.

free play cross-parent attachment emotion regulation context mother-infant father-infant

Over the past several years, the importance of emotion regulation has gained recognition in the literature on early socioemotional development (Bridges & Grolnick, 1995; Fox, 1994; Kopp, 1989). It has been suggested that early emotion regulation may be influenced by qualitative aspects of infant-caregiver interactions (e.g. Bridges & Grolnick, 1995; Cassidy, 1994; Tronick, 1989). To date, however, research has focused exclusively on emotion and emotion regulation in infant-mother interactions. The purpose of this article is to expand this focus by exploring the similarities and differences in infant strategy use and expressed emotion in differing contexts with both mothers and fathers.

Within the literature on early emotional development, the primary focus has been on developmental trends and individual differences in emotional expression. Emotion regulation researchers have expanded this focus by examining the behavioral strategies used to regulate

Direct all correspondence to: Lisa J. Bridges, 3900 Edmunds St., NW, No.2. Washington, DC 20007.

_~

emotion, as well as the associations between expressed emotion and strategy use (e.g., Braun- gart & Stifter, 1991; Eisenberg & Faber, 1992; Gianino & Tronick, 1988; Grolnick, Bridges, & Connell, 1996; Stifter & Braungart, 1995). Braungart and Stifter, for example, found that distressed 12-month-olds tended to orient toward objects during reunion episodes of the Ainsworth and Wittig (1969) strange situation, rather than orienting toward their mothers, exploring toys, or self-soothing. They theorized that temporarily focusing attention away from the source of distress (mother) to more neutral objects aided in reducing negative arousal.

Other researchers have begun to examine the role that psychological context plays in emotion regulation (e.g., Grolnick et al., 1996). A sys- tematic focus on contexts is consistent with the functionalist perspective on emotion (J. J. Cam- pos, R. G. Campos, & Barrett, 1989; J. J. Cam- pos, Mumme, Kermoian, & R. G. Campos, 1994), and with a more general trend among developmental researchers exploring diverse issues ranging from cultural influences on cog- nition (Gauvain, 1995) to stress and coping in childhood (e.g., Bridges & Grolnick, 1995; Weisz & Dennig, 1993).

47

48 Bridges, Grolnick and Connell

Despite increasing awareness of the importance of context, however, there are few pub- lished reports that directly examine the influence of variations in psychological context on the organization of infant emotion and behavior. In one recent study involving assessments of 24-month-olds and their mothers, Grolnick et al. (1996) reported strong context influences on emotion regulation strategies across four mildly to moderately stressful contexts-two separations from the mother and two frustrating delays with the mother present. They reported that play or active engagement with the environment was most common in situations where an adult was available to interact with the child, whereas comfort-seeking behavior was most common in a delay situation in which the mother was uninvolved with the toddler.

Recognition of the important role that fathers may play in children’s emotional development has increased in recent years (e.g., Bridges, Connell, & Belsky, 1988; MacDonald & Parke, 1984; Parke, 1996). Little research exists, however, on emotion regulation in infant-father interactions. In one relevant study, Bridges and Connell (199 1) examined cross-parent consistencies in infant emotion and behavior in two contexts: a free play and the strange situation. The behaviors examined were attachment-relevant social interactive behaviors described by Ainsworth, Blehar, Waters, and Wall (1978)- proximity- and contact-seeking, contact maintaining, playful distance interaction, resistance, and avoidance. All of these may be seen as strategies for regulating emotion, although this perspective has appeared infrequently in the research literature. Cross-parent consistency in emotion was found for the stressful strange situation episodes, but not for free play. Cross-parent consistency was also found in some social interactive behaviors, most notably, contact- seeking and maintaining during the strange situation.

This study was designed to expand the literature on early emotion regulation by examining consistencies and inconsistencies in expressed emotion and strategy use across contexts with mothers and fathers. The contexts used in this study were two delay situations during which the infant was required to wait to obtain an attractive object. The primary contextual varia- tion was the extent to which the parent actively

interacted with the infant. During one delay, the parent was told to interact freely, but in the other delay, the parent was asked to work on a set of questionnaires. Such a task requires the parent to focus most of his or her attention away from the infant’s ongoing activities.

Our first set of hypotheses involved differences in infant emotion and strategy use across parents and across contexts. It was expected that there would be context effects on the level of negative emotion exhibited by infants. Specifi- cally, infants were expected to exhibit more distress during delays in which the parent was preoccupied (and therefore unavailable to support the infant’s emotion regulation efforts) than during delays in which the parent was free to interact. Furthermore, we expected that there would be context influences on strategy use. Based on the Grolnick et al.‘s (1996) findings, strategies involving active engagement with some aspect of the environment (other than the delay object) were expected to be more preva- lent when the parent was interactive than when the parent was preoccupied, whereas the oppo- site was expected to hold for other behaviors, such as seeking comfort or caregiving.

Bridges and Connell (199 1) did not find differences in the mean levels of emotion expressed with mothers and fathers in the strange situation, and few differences in behaviors. No previous studies have included both mothers and fathers in contexts such as were used in this study, however, and no studies examining infant behavior with both parents have focused on behavioral strategies for emotion regulation. Therefore, mother versus father presence was included in all analyses examining context effects on infant emotion and strategy use, although no specific hypotheses were made.

A second set of hypotheses involved consistencies in individual differences in expressed emotion and strategy use cross-context and cross-parent. Despite possible mean-level differences across the two different contexts, we expected that infants who were relatively more negative in one context would be relatively more negative in the other context as well, and that strategy use would also show cross-context consistency. Such consistency has been reported previously for older toddlers (Grolnick et al., 1996). Although similar contexts have not been used in investigations of infant emotion regula-

Emotional Regulation with Mothers and Fathers 49

tion with fathers, Bridges and Connell (199 1) did report that emotion in free play with the father was significantly correlated with emotion in episodes of the strange situation. Thus, cross-context consistency was expected with both parents.

As has been found previously (Bridges & Connell, 1991; Bridges et al., 1988), we expected cross-parent consistency in the extent to which infants expressed negative emotion, and in the types of strategies displayed. This was expected particularly across parent-passive delays due to the potentially reduced influence of parent behavior witbin the immediate situation. In other words, infant emotion and strategy use in situations in which parents were relatively passive were expected to reflect more consistent, infant-based tendencies.

Finally, we examined associations between infant emotion and strategy use. We expected that focusing on the unobtainable delay object would be associated with higher levels of distress with both mothers and fathers. In contrast, we expected that active engagement with other aspects of the environment would be associated with less negative emotion in both infant-parent assessments.

More speculatively, we explored the possibility that some emotion regulation strategies would be more closely linked with negative emotion in interactions with one parent versus the other. For example, contact-seeking and requests for caregiving are described as being characteristic of infant-mother interactions in general, more so than in infant-father interactions (Parke, 1996). It may be, therefore, that this type of behavior would be less highly associated with negative emotion in infant-mother interactions than in infant-father interactions, where only children who are experiencing distress would be expected to engage in requests for caregiving.

METHOD

Subjects

The data come from one wave of a 2-year longitudinal study of emotion regulation in infancy. Two assessments are included. one with each parent. at infant ages 12% and 14 months CM = 53.97 and 59.39 weeks, SD = 2.67 and 3.04 weeks. respectively). A total of 64 infants (3 I boys, 33 girl<) were seen. Of these, 5 completed only one visit (3 with the mother, 2 with the father).

Mothers ranged in age from I7 to 44 when their infants were born (M = 29.49. SL) = 5.01); fathers ranged in age from 19 to 42 CM = 31.53, Sf) = 4.85). All families were

intact and were primarily Euro-American. One or both parents were Latin0 m eight families, and two families were Asian. Twenty-two percent of mothers and 20% of fathers had high school education. 36% of mother\ and 41’1 of fathers had post-high school training (e.g.. associate’\ 01 technical degrees), 33% of both mothers and father5 had bachelor’s degrees, and 9% of mothers and 6% of fathers had graduate training. Using the Nakao and Tress (1992: reprinted in Entwisle & Astonc, 1994) Socioeconomic Index of Occupations, occupational prestige ranged widely, from 29.03 to 96.04 (on a O-100 scale: M of 59.85. SD = 16.95). Thirty-nine mothers were employed outside the home. 30 of them full-time.

Overview of Laboratory Procedures

The order of infant-mother and infant-father viaits was counter balanced. All procedures took place in a large laboratory playroom containing a table. two desk chairs, two padded chairs. and a 5 ft (I .S2 in) high cabinet in one corner. A different female experimenter conducted each of the two visits for each child. All procedures were videotaped through a one-way mirror for later coding.

Each visit involved two 6-min delay episodes, separated by IO min of free play. During each delay. the infant was provided with a few moderately interesting toys (e.g.. a shape sorter, a soft ball. a small teddy bear). For one cpisodc, the parent was asked to refrain from interacting with the infant (the /““r~~i[-l”“,~i~‘r r/&v). No such limitations were placed on the parent durmg the other (/~trren/-trc.li~,~) delay. The order of the delays and the two stimuli (food, gift) prc- cented during each delay were counterbalanced.

Thi\ delay wa\ adapted from a procedure developed by J. H. Block and J. Block ( 1980) and later used by Grolnick ct al. (1996). The experimenter brought a toy. wrapped in clear cellophane with bright ribbons and stickers. into the room and let the child look closely at it. She then said that she had forgotten something outside and that hhe would give the infant the gift when she returned. She placed the gift on the cabinet. out of reach but within the infant’s sight, and left the room for 6 min. The episode ended when she returned and gave the gift to the infant.

This procedure, also used by Grolnick et al. (1996) wah the \ame as the gift delay except that instead of a gift, the delay object was a small cup of animal crackers. One other modi- fication was that the experimenter entered the room after 3 minn, gave the child one cracker. and left again. This break was included to maintain consistency with procedures estab- lished by Grolnick et al. The decision to include a break in that study was the result of pilot work conducted by Grolnick et al. in which a break was found to be necessary to equal& the degree of interest that toddlers demonstrated in the two delay objects.

These two conditions were included in order to asses expressed emotion and strategy use under varying levels of external support and assistance. In the parent-passive condi- tion, the parent was asked to sit at a table and to work on questionnaires. The parent was told not to initiate interaction


but that he or she should respond to the infant’s initiations if necessary. In the parent-active delay, the parent was asked to interact freely with the child.

Observational Coding of Infant Behaviors

Coding was conducted in 5-s intervals across each 6-min episode. A total of six coders participated: Two conducted strategy coding exclusively, two coded emotion exclusively, and two participated in both types of coding.

Emotion

Facial and vocal emotion scales developed by Thompson and Lamb (1984) were used. The facial scale ranged from brighf smile (1) tofull cryface (8). If the child’s face was not visible during an interval, no code was given. The vocal scale ranged from intense delight (1) to hyperventilated cry (14). If the child did not vocalize during an interval, a neutral code (4) was given. In all cases, the code given in an interval was based on the most extreme (i.e., furthest from neutral, either positive or negative) expression observed.

To create summary scores, facial and vocal emotion codes were first averaged within intervals. Scores describing temporal and intensity features of emotion within each episode were then created. These included: a) the mean emotion rating; b) the peak, or most negative, emotion rating reached; c) the latency (number of intervals) to initial onset of distress; d) the number of bouts of negative emotion; and e) the mean duration of negative emotion bouts. A bout was defined as one or more intervals in which the child displayed negative emotion (a code above 4.0). preceded and followed by one or more intervals of neutral or positive emotion (codes of 4.0 or below). For each bout, duration was the number of consecutive intervals during which negative emotion was expressed.

Table 1 presents the strategy codes. The same codes were used for all episodes. Infants could receive up to three different codes within an interval. However, due to the extremely high frequency with which behaviors denoting passive engagement could be coded (e.g., because the child might glance briefly at something in the room while actively engaged in play), this strategy was coded only when no other behavior was seen in the interval. This decision was made at the time the coding system was being developed, and was motivated by concerns about conceptual clarity and useful- ness, and by problems with interrater agreement.

Scores reflecting frequencies for each strategy in each episode were created. Initial examinations revealed posi- tively skewed distributions for a number of the strategies. In the parent-active situation, all strategies with the exceptions of active engagement with the parent and parent-related passive engagement were skewed. In the parent-passive situation, all strategies except active engagement, alone, and passive engagement were skewed. Because of this square root transformations were performed on these variables, and transformed scores were used for all analyses.

One other data transformation was made: In parent-passive episodes, only half of the children attempted to engage the mother in play, and a similar number attempted to engage the father. Because of this, dichotomous variables were created reflecting whether the child did, or did not, use this strategy.

TABLE 1 Summary of Emotion Regulation Strategies

1.

2.

3.

4.

5.

Active Engogemenf. Ploy or exploratory activities that involve on active engagement with some aspect of the environment.

1.1 A/one. Using toys for intended purposes, other task-directed activities; sustoined, intense visual examination of on object.

1.2 With the parent /Attempts to engage parent. Any of the previously-mentioned activities when performed with the parent; includes ottempts to engoge the porent and ongoing reciprocal octiviiy.

Passive Engagement. Fairly possive use of the environment, when the child’s behavior is not actively task directed. This includes both nonsocial (e.g., unsustained, cosuol interest in toys or objects in the room, wondering around the room) ond social (e.g., attending to toys or objects held by the porent, responding briefly to the parent) behavior.

Physical Self-soothing. Mouthing honds, fingers, or clothing. Hugging, mouthing, or lying on o toy. Lying on the floor, covering eyes, rocking, or rubbing ports of the body (wringing honds, rubbing heod, pulling ears, twisting hair, etc.).

Other-directed/Focus on the Porent. Proximol or distol comfort-seeking behoviors, requests for core- taking ossistonce, ond behoviors indicating on interest in being with the parent OS on end unto itself.

Focus on the Delay Object. Looking at, pointing to, or repeating the nome of the object. Calling for porentol assistonce and active attempts to retrieve the object. Colling or otherwise searching for the experimenter.

Interrater reliabilities for strategy coding were assessed using Cohen’s kappas. For each of the four situations (mother-active, mother-passive, father-active, father-passive), kappas were based on randomly chosen subsamples of 15 subjects who were coded twice. Kappas were .8 1 and .76 for strategy use in parent-active situations with mothers and fathers, respectively. Kappas were .88 and .90 for mother- passive and father-passive situations, respectively.

lntraclass correlations were computed in order to assess interrater agreement for facial and vocal emotion variables. With the mother, the four correlations ranged from a low of .82 for facial emotion in the mother-active episode to a high of .98 for facial emotion in the mother-passive episode. With the father, correlations ranged from .93 for vocal emotion in the father-active episode to .99 for facial emotion in the father-passive episode.

Results

Preliminary Analyses

Composite emotion variables were created based on the results of principal components factor analyses, conducted separately for mother-passive, mother-active, father-passive,


and father-active episodes. In all cases, one factor was indicated based on the mine&en crite- rion. Eigenvalues of the four factors ranged from 3.56 to 3.75. All factor loadings were greater than .77. Based on these results, four negative emotion composites were created by averaging the standardized scores for the five emotion indices for each episode.

Context and Parent Injluences on Infant Emotion and Strategy Use

We expected that there would be contextual influences on the levels of emotion expressed by infants, and that these influences would be similar in infant-mother and infant-father assessments. To examine this, repeated measures ANOVAs for negative emotion and for each strategy variable were computed with context (parent-active vs. parent-passive) and parent (mother vs. father) as within-subject factors. Sex of child was included as a between-subjects factor.

There was a main effect of context for emotion, F(1, 55) = 12.47, p<.OOl. Infants were more negative in parent-passive than in parent- active contexts. There was also a marginal effect of parent, F(1, 55) = 3.59, p x.10, with infants tending to express more negative emotion with the mother than with the father. There were no significant sex of child effects or interactions, Means of the composite emotion scores were as follows: mother-active, M = -.07, (SD = .52); father-active, M = -.17 (SD =.35); mother-passive, M = .21 (SD = .66); and father-passive, M = .04 (SD = .63). Again, lower scores reflect less negative emotion.

Analyses addressing the influences of context and parent on strategy use were conducted next. For active engagement, three repeated measures ANOVAs were conducted. The first compared overall active engagement, which was a combined measure of active engagement alone and with the parent. This was done to assess comparability of our results with those of Grolnick et al. (1996). The second ANOVA examined active engagement alone. A third repeated measures ANOVA with parent as the single within-subjects variable was conducted for active engagement with the parent during the parent-active episodes.

Context effects were found for some strategy variables. Means of the square root transformed

data for the parent-active and parent-passive episodes are reported in Table 2. Active engagement, overall, was significantly more common in parent-active than in parent-passive contexts, although active engagement alone was more common in parent-passive episodes. Continued focus on the delay object was more likely in parent-passive than in parent-active situations, as was focus on the parent. Physical self-soothing behaviors were also marginally more frequent in the parent-passive than in the parent-active situations.

No main effects of parent were found for strategy variables. There was one main effect of

TABLE 2 Means of Square Root Tranformed Data,

Infant Strategy Use in Parent-Active and Parent-Passive Delay Situations

with Mothers and Fathers

Parent- Parent- Strategy Active Delay Passive Delay F Variable M (SD) M (SD) (1, 55)

1. Active Engagement

1 a. Total (Alone ond with parent)

Over011 5.72 (1.33) 4.30

With 5.71 (1.60) 4.04

Mother

With Father 5.78 (1.65) 4.43

1 b. Alone

Overall 2.57 (1.30) 3.96

With 2.23 (1.65) 3.68

Mother

With Father 2.86 (2.17) 4.04

2. Passive Engagement

Overall 4.51 (1.23) 4.64

With 4.45 (1.46) 4.56

Mother

With Father 4.60 (1.52) 4.65 3. Physical Self-Soothing

Overall 1.94 (1.23) 2.36

With 2.17 (1.62) 2.55

Mother

With Father 1.86 (1.68) 2.32

4. Focus on the Parent

Overall 1.71 (1.59) 2.83

With 1.81 (2.00) 2.92

Mother

With Father 1.57 (2.04) 2.68 5. Focus on the Delay Object

Overall 1.84 (1.00) 2.32

With 1.90 (1.36) 2.44

Mother

(1.83) 31.95”

(2.45)

(2.21)

(1.81) 26.87”

(2.54)

(2.17)

(1.25) .62

(1.85)

(1.60)

(1.48) 3.46+

(2.07)

(2.11)

(1.60) 18.35’*

(2.17)

(1.93)

( .86) 13.11”

(1.15)

WithFather I.77 (1.12) 2.22 (1.16)

Note. +p<.lO, ‘p<.OOl, **p<.OOOl


child sex. Boys demonstrated more active engagement alone than did girls, F( 1,55) = 5.52, p <.05. Univariate ANOVAs revealed that this difference, although consistently in the same direction across all episodes, was significant only in the father-active episode (M = 3.58, SD = 2.02 for boys, and M = 2.14, SD = 2.10 for girls,F(1,55)=7.15,p<.Ol).

There was one significant parent x child sex interaction. Boys displayed more physical self- soothing with their mothers than with their fathers, whereas for girls, physical self-soothing did not differ cross-parent, F( 1, 55) = 4.35, p c.05. Boys’ means were 2.12 (SD = 1.70) and 2.82 (SD = 2.48) in mother-active and mother- passive episodes, respectively, whereas they were 1.49 (SD = 1.49) and 1.99 (SD = 2.18) in the father-active and father-passive episodes. For girls, the four means (in the same order) were 2.23 (SD = 1.57), 2.29 (SD = 1.57), 2.22 (SD = 1.80), and 2.63 (SD = 2.02).

In the repeated measures ANOVA assessing parent and child sex effects on active engagement with the parent during parent-active episodes, no main effects were found. There was, however, a significant parent x child sex interaction F(1, 55) = 4.02, p c.05. Boys engaged in more active play with the mother (M = 5.06, SD = 1.93) than with the father (M = 3.93, SD = 2.18). Girls did not differ in their active engagement with mothers (M = 4.74, SD = 1.85) and fathers (M = 4.82, SD = 1.87).

Cross-Context Consistency of Individual D@erences

These data support the expectation that infants would exhibit more negative emotion during parent-passive episodes than during parent- active episodes. We also hypothesized that there would be some intraindividual consistency in infants’ tendencies to exhibit negative emotion across the two contexts. Contrary to expectations, however, negative emotion scores were not significantly correlated cross-context with the mother, 462) = .07, ns. In the infant-father assessment, individual differences in infant negative emotion expression did show low but significant consistency, r(60) = .29, p c.05. A test of the significance of the difference between these two dependent correlations revealed that the difference was marginally significant, z = 1.27,pc.lO.

Examinations of cross-context correlations for emotion regulation strategies indicated significant correlations for three of the six strategies in the infant-mother assessment: active engagement, alone, r(62) = .27, p x.05., passive engagement, r(62) = .26, p c.05, and focus on the delay object, r(62) = .38, p x.01. None of the cross-context correlations for the infant-father assessment was significant. Correlations corre- sponding to the three significant infant-mother assessment correlations were .20 (ns). .23 @ <. lo), and .21, ns, respectively. These correlations, although not significant, did not differ in magnitude from infant-mother assessment correlations.

Cross-Parent Consistency in Infant Emotion and Strategy Use

We hypothesized that there would be individual consistency across the two assessments, particularly across parent-passive episodes. Results for infant emotion were consistent with this expectation. Negative emotion was not significantly correlated across parent-active episodes, 457) =. 10, ns, whereas there was a moderate, significant correlation across parent-passive episodes, r(58) =.35, p c.01. A test of the significance of the difference between dependent correlations indicated that there was a marginally significant difference between them, z = 1.44, p <. 10.

Cross-parent correlations for strategy use were again examined separately for parent- active and parent-passive episodes. Little consistency was found. Across the mother-active and father-active episodes, passive engagement was significantly correlated, r(57) = .4 1, p x.0 1, and there were marginally significant correlations for focusing on the parent for caregiving or comfort r(57) = .23, p x.10 and focusing on the delay object, 457) = .22, p <. 10. Across parent-passive episodes, the only significant association involved the dichotomous variable ta ping active attempts to get the parent to play x P (l,N= 58) =7.53,p<.Ol.

Relations Between Emotionality and Strategy Use

Finally, we examined associations between emotion and strategy use within-situation. We expected to find significant associations between levels of emotion exhibited and the frequencies with which different strategies were


employed. Furthermore, we speculated that some differences in the strength of the associations between emotion and strategy use would exist between infant-mother and infant-father assessments. Within context, within-parent correlations between strategy use and negative emotion are presented in Table 3.

As can be seen from Table 3, three strategies demonstrated significant or marginal correlations with emotion in all episodes: a) active engagement alone was associated with less negative emotion; b) focus on the delay object was associated with more negative emotion; and c) focus on the parent for comfort or caregiving was associated with more negative emotion, although the correlation was marginally significant in the mother-active episode. Associations between emotion and other strategy variables were less consistent across episodes.

Results presented in Table 3 also suggest that associations between strategy use and emotion differed between the mother-active episode and the other three episodes in two cases. First, the association between focusing on the delay object and emotion was stronger in the mother- active delay. Tests of the significance of the difference between dependent correlations indicated that the mother-active delay correlation was significantly stronger than correlations in both the mother-passive and father-passive delays (z = 1.99, and z = 1.8 1, respectively ps <.05), and was marginally stronger than the

TABLE 3 Associations Between Negative Emotion

and Strategy Use

ESR Strategy

Episode

MotherMotherFctther-Father- -Active -zz:- Active Passive

* Active Engagement, -.3 1 -.41’” -.32* -.33- Alone

Active Engagement, -.33” -.02 -.14 -.23+

With Parent Passive Engagement .06 -.Ol -.12 -.13

Physical Self- .43*** .25+ .27* .16

Soothing

FOCUS on the Parent .24+ ,5,“” ,51 **It ,47.‘“.

Focus on the Delay .56’*” .27’ .33” .29’ Obieci

Note. +p<.10, ‘pCI.05, **p<.o1, “*p~.ool,

-**p<.0001.

father-active delay correlation (z = I .26, p <. 10). Second, focusing on the parent for comfort or caregiving was less strongly associated with emotion in the mother-active delay than during the other episodes. Again, tests of the significance of the differences between dependent correlations revealed that the mother-active delay correlation was significantly lower than the correlations in the mother-passive and father-active delays (z=2.25 and z = 1.72, respectively, ps c.05, and was marginally lower than the father- passive delay correlation (z = 1.42, p c.10).

DISCUSSION

This study was designed to examine infant emotion regulation across two contexts in the presence of the mother and of the father. Consis- tent with expectations, we found significant influences of context on emotion. Infants were significantly more negative (and less positive) during situations in which the parent performed activities other than interacting with the infant, than in situations where the parent was actively engaged with the infant.

As expected, we also found context influences on emotion regulation strategy use. Active engagement was more frequent in parent-active than in parent-passive episodes, although solitary active engagement was higher in the parent- passive episode. This is consistent with the findings of Grolnick et al. (1996) for 24-month-olds. As noted earlier, infants and young children may be most able to maintain a focus on playlike activities when interacting with caregivers (Bridges & Grolnick, 1995; Kopp, 1989). When the parent is unavailable or uninvolved, this reduces the ability of the child to use active strategies for emotion regulation. Parental unavailability also may become an additional stressor, leading to more negative emotion and a reduced ability to distract attention away from the original stressor (as shown by higher frequencies of focus on the delay object) or from the additional stressor (indicated by increased focus on the parent for contact and caregiving).

There was little evidence of differences in infant emotion or strategy use in contexts with fathers versus mothers. There was some indica- tion, however, that boys’ behavior may be more influenced by whether the parent present is the mother or the father. This was suggested by the significant parent by child sex interactions for


physical self-soothing behaviors and for active engagement with the parent during parent-active episodes, as well as the significant sex of child effect for solitary active engagement during the parent-active contexts (which univariate tests revealed was carried primarily by differences in the father-active delay).

Our data do not allow us to assess the possible causes of the apparently more differential use by boys of particular emotion regulation strategies. One possibility is that boys are more strongly influenced than are girls by differences in maternal and paternal interactive styles, such as the reported tendencies of fathers to engage in more parallel play with infants than do mothers (Lamb, 1977), and to be more directive and intrusive in play interactions than are mothers (Power, 1985). Greater sensitivity to mother- father differences could potentially be associated with boys’ tendency to be less actively engaged with the father than with the mother, and to exhibit more solitary active engagement than do girls in father-active episodes.

A second possibility is that fathers treat boys and girls more differentially than do mothers. It has been reported that fathers encourage sex- typed behavior more strongly than do mothers, and that this is particularly true for fathers of boys (Langlois & Downs, 1980). Fathers of boys may discourage some physical self-soothing behaviors, such as hugging a stuffed animal, whereas fathers of daughters may be less inclined to do so. Mothers, in contrast, may react similarly to their sons’ and daughters’ physical self-soothing behaviors.

We expected to find consistent individual differences in infant emotion both cross-parent and cross-context. In contrast to the significant associations between toddler emotion in mother-passive and mother-active delays reported by Grolnick et al. (1996), infant emotion with the mother was not consistent cross- context in this investigation. There was a significant cross-context association with the father, however. This greater association with the father than with the mother was consistent with the findings of Bridges and Connell (1991).

This is the first study we are aware of that specifically examines cross-context consistencies in emotion regulation strategies among infants of this age. Consistent with Grolnick et al. (1996), infants who continued to focus on an

object that was present, but unavailable, in the mother-active context were likely to do so in the mother-passive context as well. Further- more, both active and passive engagement with the environment demonstrated cross-context consistency with the mother. No cross-context correlations were significant with the father. These results do not, however, suggest that infant behavior was more consistent with the mother than with the father, because the direction and magnitude of two of these three correlations were similar with the two parents.

Cross-parent, we expected to find consistency in infant emotion and strategy use. Given the potentially reduced influence of direct parental intervention in parent-passive episodes, we particularly expected consistency across those episodes. Our findings for emotion were consistent with these expectations, suggesting that when the direct involvement of the parent in regulating the infant is reduced, consistencies in infant emotion become more evident. Given the significantly more negative level of emotion expressed in parent-passive than in parent- active contexts, the findings may also be inter- preted as consistent with those of Bridges and Connell (1991). They found that cross-parent consistency in emotion was relatively higher across contexts that pulled for more negative emotion (episodes of the strange situation) than across contexts where emotion was generally more neutral or positive (free play).

Contrary to expectations, there was little cross-parent consistency in the strategies examined in this study. In parent-active situations, the only behavior that was significantly correlated cross-parent was passive engagement. Because this strategy was coded only when no other behavior was evident, this correlation may indi- cate a tendency to exhibit this behavior versus doing anything else. It is possible that this type of passive engagement may reflect temperament characteristics of the infant, such as low activity level.

In the parent-passive situation, the only behavior that was consistent cross-parent was attempting to engage the parent. Only about half of the infants exhibited this behavior at all with either parent, and the frequency with which infants did so varied tremendously. It is of interest, however, that infants who engaged in this behavior in at least one 5-s interval with one par-


ent were significantly more likely to do so with the other parent as well. Initiation of playful interchange, particularly during a mildly stressful situation, may represent a fairly sophisti- cated strategy for these infants. If so, assessments at later ages may reveal higher base rates and stronger, more consistent individual differences.

Taken together, the cross-context and cross- parent results may suggest that, although stable individual differences in emotion expression and emotion regulation strategy use may be beginning to emerge in later infancy, infant emotional behavior is strongly influenced by characteristics of the immediate context. Such characteristics include which caregiver is present, and the extent to which that caregiver is interactive with the infant. As noted previously, it is possible that individual differences in strategy use may become more consistent cross-caregiver, as well as cross-context as the child develops.

Finally, we hypothesized that emotion would be associated with the types of strategies employed by infants. As expected, infants who displayed relatively little active engagement and more frequent focus on the delay object showed more negative emotion in both contexts and with both parents. These results are consistent with Grolnick et al. (1996). They may also be seen as consistent with the literature on coping among older children and adults (e.g., Folkman & Lazarus, 1990; Guskin, 1993). Within the coping literature, studies have found that rumi- nation or vigilance in uncontrollable situations is associated with more negative emotion. Active deployment of attention and self-distraction, in contrast, have both been associated with more positive emotion.

It is interesting to note that, although the overall pattern of correlations between emotion and strategy use was similar in different contexts and with the mother and father, there were two notable variations. Specifically, the association between focus on the delay object and negative emotion was stronger, and the association between focus on the parent and negative emotion was weaker in the mother-active episode than in other episodes. With respect to focus on the parent, it is possible that close physical contact, as well as requests for, and receipt of, caregiving are integral components of even playful

interactions between mothers and infants, more so than between fathers and infants. This inter- pretation is consistent with descriptions of differences between maternal and paternal interactions with infants (Parke, 1996). This could reduce the association between negative emotion and this strategy in the mother-active episode, relative to associations found in other episodes.

In summary, the results of this study were fairly consistent with previous reports indicating some consistencies in individual differences in infant emotion and in behaviors that may be conceptualized as strategies for emotion regulation, as well as large effects of context on emotion and strategy use. Furthermore, this study indicates that infants, particularly boys, may be beginning to develop different behavioral reper- toires for interactions with different caregivers, perhaps based on differential characteristics of these relationships.

There are a number of limitations to this report that should be addressed in subsequent studies. First, the data presented are cross-sec- tional. Additional longitudinal analyses will be necessary to examine the extent to which consistencies in infant emotion and behavior, and early differential characteristics of infant-mother and infant-father interactions as they appear to impact infant emotion regulation, will influence the continuing development of styles of emotion regulation beyond infancy.

Finally, contextual influences on infant emotion regulation were a central focus of this study. Contextual influences have often been simply assumed or entirely ignored, as sources of pre- dictable variability in infant behavior and development. Yet, when context is ignored, infant emotion and behavior may be seen as being either more consistent than they actually are (e.g., when a single context is included in a study), or as less consistent (e.g., when multiple situations are included without consideration of differences in psychological supports for, and impacts on, the infant).

It was our intention in this study to include two contexts that varied along one primary psychological dimension-the degree to which the parent was involved with, versus disengaged from, interaction with the infant. This dimension should be particularly relevant to examinations of emotional self-regulation during a develop-


mental period in which regulation is thought to become increasingly internal (Bridges & Grolnick, 199.5; Kopp, 1989). This study is admittedly a first step. Previous studies have included more diverse contexts (Grolnick et al., 1996), but have not included fathers as well as mothers. A more thorough understanding of the importance of contextual influences on emotion and emotion regulation strategy use will be facilitated by additional studies that systemati- cally vary the contexts in which infants and their parents are observed, along other psychological dimensions that are theoretically relevant to the development of emotional self-regulation.

ACKNOWLEDGMENTS

This work was supported in part by grant MH44449 from the National Institute of Mental Health. We would like to thank Michael Morales and Sherri Addis Palmer for their assistance in data coding.

Further information on the development of laboratory procedures can be obtained from W. Grolnick, Frances L. Hiatt School of Psychol- ogy, Clark University, 950 Main St., Worcester, MA 01610.

REFERENCES

Ainsworth, M.D..%, Blehar, M.D., Waters, E., & Wall, S. (1978). Patterns ofattachment: A psychological study ofthe strange situation. Hills- dale, NJ: Erlbaum.

Ainsworth, M.D.S., & Wittig, M. (1969). Attachment and exploratory behavior of one-year-olds in a strange situation. In B.M. Foss (Ed.), Determi- nants of infant behavior Vol. 4. London: Meth- uen.

Block, J.H., & Block, J. (1980). The role of ego con- trol and ego resiliency in the organization of behavior. In W.A. Collins (Ed.), Minnesota symposia on child psychology, Vol. 13. Hills- dale, NJ: Erlbaum.

Braungart, J.M., & Stifter, C.A. (1991). Regulation of negative reactivity during the strange situation: Temperament and attachment in 12-month-old infants. Infant Behavior and Development, 14, 349-364.

Bridges, L.J., Connell, J.P., (1991). Consistency and inconsistency in infant emotional and social interactive behavior across contexts and caregivers. Infant Behavior and Development, 14, 471-487.

Bridges, L.J., Connell, J.P., & Belsky, J. (1988). Sim- ilarities and differences in infant-mother and

infant-father interaction in the strange situation: A component process analysis. Develop- mental Psychology, 24,92-100.

Bridges, L.J., & Grolnick, W.S. (1995). The development of emotional self regulation in infancy and early childhood. In N. Eisenberg (Ed.), Review of personality and social psychology: Vol. 15. Social development. Thousand Oaks, CA: Sage.

Campos, J.J., Campos, R.G., & Barrett, KC. (1989). Emergent themes in the study of emotional development and emotion regulation. Develop- mental Psychology, 25, 394-402.

Campos, J.J., Mumme, D.L., Kermoian, R., & Cam- pos, R.G. (1994). Commentary: A functionalist perspective on the nature of emotion. In N. Fox (Ed.), The development of emotion regulation: Biological and behavioral considerations. Monographs of the Society for Research in Child Development, 59 (2-3, Serial No. 240) 284-303.

Cassidy, J. (1994). Emotion regulation: Influences of attachment relationships. In N. Fox (Ed.), The development of emotion regulation: Biological and behavioral considerations. Monographs of the Society for Research in Child Develop- ment, 59, (2-3, Serial No. 240) 228-249.

Eisenberg, N., & Fabes, R.A. (1992). Emotion, regulation, and the development of social competence. In M.S. Clark (Ed.), Review of personality and social psychology, Vol. 14: Emotion and social behavior. Newbury Park, CA: Sage.

Entwisle, D.R., & Astone, N.M. (1994). Some practi- cal guidelines for measuring youth’s race/eth- nicity and socioeconomic status. Child Development, 65, 1521-1540.

Folkman, S., & Lazarus, R.S. (1990). Coping and emotion. In N.L. Stein, B.L. Leventhal, & T. Trabasso (Eds.), Psychological and biological upproaches to emotion. Hillsdale, NJ: Erlbaum.

Fox, N.A. (Ed.) (1994). The development of emotion regulation: Biological and behavioral considerations. [Special Issue] Monographs of the Soci- ety for Research in Child Development, 59, (2- 3, Serial No. 240).

Gauvain, M. (1995). Thinking in niches: Sociocul- tural influences on cognitive development. Human Development, 3S25-45.

Gianino, A., & Tronick, E.Z. (1988). The mutual regulation model: The infant’s self and interactive regulation, coping, and defense. In T. Field, P. McCabe, & N. Schneiderman (Eds.), Stress and coping. Hillsdale, NJ: Erlbaum.

Grolnick, W.S., Bridges, L.J., & Connell, J.P. (1996) Emotion regulation in two-year-olds: Strate-


gies and emotional expression in four contexts. Child Development, 67, 928-941.

Guskin, K.A. (1993, March). Coping with uncontrollable stressors: Children’s know’ledge qf rwni- nation and distraction. Paper presented at the biennial meeting of the Society for Research in Child Development, New Orleans, LA.

Kopp, C.B. (1989). Regulation of distress and negative emotions: A developmental view. Devel- opmental Psychology, 25, 343-354.

Lamb, M.E. (1977). Father-infant and mother-infant interaction in the first year of life. Child Devel- opment, 48, 167-181.

Langlois. J.H., & Downs, A.C. (1980). Mothers, fathers, and peers as socialization agents of sex-typed play behavior in young children. Child Development, 51, 1237-1247.

MacDonald, K., & Parke, R.D. (1984). Bridging the gap: Parentshild play interaction and peer interactive competence. Child Developmenr, 55, 1265-1277.

Parke, R.D. (1996). Fatherhood. Cambridge, MA: Harvard University Press.

Power, T.G. (1985). Mother- and father-infant play: A developmental analysis. Child Development, 56, 1514-1524.

Stifter, C.A., & Braungart, J.M. ( 1995). The regulation of negative reactivity in infancy: Function and development. Developmenral Psychology. 31,448a55.

Thompson, R.A., & Lamb. M.E. ( 1984). Assesing qualitative dimensions of emotional respon- siveness in infants: Separation reactions in the strange situation. Infant Behavior and Devel- opment, 7,423-I45.

Tronick, E.Z. (1989). Emotions and emotional com- munication in infants. American Psychologisf 44, 112-l 19.

Weisz, J.R., & Dennig, M.D. (1993, March). The search for an understanding of “good” stress and coping in childhood. Paper presented at the biennial meeting of the Society for Research in Child Development, New Orleans, LA.

30 August 1995; Revised 20 June 1996 n

Documents

Infant emotion regulation with mothers and fathers