Abstract
There is considerable knowledge of parental socialization processes that directly and indirectly influence the development of children’s emotion self-regulation, but little understanding of the specific beliefs and values that underlie parents’ socialization approaches. This study examined multiple aspects of parents’ self-reported emotion representations and their associations with parents’ strategies for managing children’s negative emotions and children’s emotion self-regulatory behaviors. The sample consisted of 73 mothers of 4–5-year-old children; the sample was ethnically diverse. Two aspects of parents’ beliefs about emotion – the importance of attention to/acceptance of emotional reactions, and the value of emotion self-regulation – were associated with both socialization strategies and children’s self-regulation. Furthermore, in mediational models, the association of parental representations with children’s emotion regulation was mediated by constructive socialization strategies. These findings are among the first to highlight the specific kinds of emotion representations that are associated with parents’ emotion socialization, and their importance to family processes shaping children’s emotional development.
Keywords
Parents socialize emotion regulation directly and indirectly. Direct approaches include coaching children’s self-regulation of emotion, while indirect efforts including managing the emotional demands of family life (Thompson & Meyer, 2007). Underlying these multifaceted strategies are the parents’ own beliefs and values about emotion. These include beliefs about the importance of paying attention to emotions, values about the consequences of expressing emotion, and emotional self-awareness (see Parker et al., 2012, for an insightful analysis). Developmental researchers have been especially interested in parental beliefs and values about emotion because of their potential importance to the intergenerational transmission of emotional response characteristics. Because parental beliefs about emotion become incorporated into specific socialization practices that influence children’s emotion self-regulation, this can help to account for how adaptive or maladaptive styles of emotion expression and regulation are conveyed from parent to child (see, for example, Baker, Fenning, & Crnic, 2011; Wong, Diener, & Isabella, 2008; Wong, McElwain, & Halberstadt, 2009). The purpose of this study was to better understand specific parental beliefs about emotion, their association with emotion socialization practices, and the direct and indirect associations of these beliefs with children’s emotion regulation.
There is considerable evidence that how parents respond to children’s emotions is important to children’s developing self-regulatory capabilities. Four-year-olds who were more emotionally competent had parents, according to one study, who were emotion coaches and maintained a positive demeanor toward the child even during parent–child conflict (Denham, Mitchell-Copeland, Strandberg, Auerbach, & Blair, 1997). In a study with older children, mothers’ self-reported problem-solving responses to their grade-school children’s negative emotions were associated with children’s constructive coping with problems, such as seeking support and positive thinking (Eisenberg, Fabes, & Murphy, 1996). These findings are consistent with many others documenting the association of parents’ supportive, constructive responses to children’s emotions with the development of more competent emotion regulation skills, and parents’ dismissing, critical, or punitive reactions with less competent emotion regulation in offspring (see Denham, Bassett, & Wyatt, 2007; Thompson & Meyer, 2007, for reviews).
Much less is known, however, about the particular parental beliefs underlying these emotion-related socialization strategies. A variety of emotion-related beliefs are potentially relevant to parental socialization efforts. These include parents’ acceptance of the validity of emotional experience, beliefs in the importance of attending to emotions, values concerning emotion self-regulation, depth of emotion understanding, and beliefs about the significance of controlling emotional displays, especially negative ones, that may be enlisted with respect to one’s children. Understanding the relevance of specific emotion-related beliefs is important to research on the origins of parental socialization efforts concerning emotion. Research studies have found, for example, that mothers who believe in the importance of guiding emotional development are more likely to talk about emotions and label emotions with their preschool children who are, in turn, more advanced in emotion understanding (Dunsmore & Karn, 2001; Perez Rivera & Dunsmore, 2011).
A significant contribution to this understanding is the concept of a parental meta-emotion philosophy, defined as “an organized set of feelings and thoughts about one’s own emotions and one’s child’s emotions” (Gottman, Katz, & Hooven, 1997, p. 243). Based on detailed coding of responses to an extended interview, Gottman and colleagues distinguish “emotion coaching” and “emotion dismissing” parenting styles on the basis of how parents regard the importance of emotions and emotional expression, the validity of emotional experience, and how parents believe they should assist the child’s emotion self-regulation. Emotion coaching parents are attentive to their own emotions and the feelings of the child, view children’s negative emotions as an opportunity for intimacy and teaching about emotions and coping, and discuss goals and strategies for managing emotions with their children. Emotion dismissing parents, by contrast, tend to ignore their own emotions or belittle their importance, do not constructively attend to their children’s feelings, and view their role as helping negative emotions change or go away.
Considerable research based on this formulation, most of it focused on emotion coaching by parents, has confirmed the predicted association of the emotion coaching style with greater emotional competence and psychosocial adjustment in children (see Katz, Maliken, & Stettler, 2012, for a review). Yap, Allen, Leve, and Katz (2008) found, for example, that mothers endorsing a more constructive meta-emotion philosophy exhibited fewer negative emotion socialization practices, and that this was moderated by the temperamental qualities of their adolescent offspring. Studying a sociodemographically at-risk sample, Cummingham, Kliewer, and Garner (2009) found that caregivers’ emotion coaching style predicted children’s better emotion understanding and emotion regulation, and was also negatively associated with later internalizing and externalizing behavior.
The research on parents’ meta-emotion philosophy has been generative, but because this work has been based on inclusive parental styles that incorporate emotion-related beliefs, child-rearing philosophy, and specific parenting strategies, it has been more difficult to identify the specific beliefs and values of parents related to their own emotions that predict their emotion socialization practices and, directly or indirectly, children’s emotion regulation. Doing so requires distinguishing between different aspects of parents’ beliefs about emotion to identify their specific associations with other aspects of parental socialization of emotion. To illustrate, one research group reported that mothers’ beliefs about the importance of attending to and accepting their own emotions predicted how insightfully they appraised the feelings of their 4-year-olds during an emotion regulation task in which they had both participated (Waters et al., 2010).
The goal of this study, therefore, was to distinguish between different emotion-related beliefs and assess their associations with mothers’ socialization of children’s emotions and children’s emotion regulation strategies. To accomplish this, we used two well-validated measures of adult representations of emotion: the Trait Meta-Mood Scale (TMMS; Salovey, Mayer, Goldman, Turvey, & Palfai, 1995) and the Emotion Regulation Questionnaire (ERQ; Gross & John, 2003). Each measure yields multiple dimensions of emotion self-awareness including acceptance of one’s feelings, clarity in emotion self-understanding, belief in the importance of emotion self-regulation, and valuing emotion inhibition or suppression. We were especially interested in four dimensions of parents’ representations of emotion. Firstly, the clarity with which parents interpret and comprehend their own emotional experience would be expected to be a significant resource to parents’ efforts to identify and understand their children’s feelings. Secondly, the extent to which parents believe that it is important to attend to and accept emotions as valid reflections of personal experience would be expected to predict how accepting they are of their children’s emotions, by contrast with dismissing them. Thirdly, how much parents strive to regulate their emotional experience by reducing negative moods and maintaining positive ones would be expected to influence how actively they seek to help their children engage in emotion self-regulation, by contrast with ignoring children’s negative feelings. Finally, and more negatively, the extent to which parents engage in emotion suppression as a personal strategy would be expected to influence how much they expect their children to inhibit negative feelings without having them recognized and validated.
We expected that mothers whose emotion representations emphasize attention to emotional experience, attention to and acceptance of one’s feelings, and the value of regulating emotions would be more likely to report problem-focused and emotion-focused strategies when responding to their children’s negative emotions, and a home environment characterized by greater positive emotional expression. By contrast, we anticipated that mothers who endorsed greater suppression of emotional expression would be more likely to report engaging in punitive or dismissing reactions to children’s negative emotions, and a more negative family emotional environment. We also expected that maternal emotion representations and reports of emotion-related socialization practices would be associated with children’s emotion regulation strategies, with mothers’ more constructive emotion representations and socialization strategies associated with children’s greater use of problem- and emotion-focused self-regulatory approaches. We were especially interested in whether mothers’ socialization strategies would mediate the influence of maternal representations on children’s emotion regulation, consistent with the conceptualization of parents’ meta-emotion philosophy. However, in light of the reporting methods of this study, such analyses must be considered exploratory.
Method
Participants
The sample consisted of 73 mothers who had preschool children between the ages of 4 and 5 years (M = 4.52 years, SD = .35; 45% female). Participants were recruited from preschools and childcare centers in three socioeconomically diverse counties. Mothers reported their own ethnicity and their child’s ethnicity: 58% of parents and 57% of children were White, 24% of parents and 19% of children were Hispanic or Latino, 8% of parents and 4% of children were Asian American or Pacific Islander, 1% of parents and children were African American, and two or more ethnicities were listed for 8% of parents and 19% of children. Maternal education was reported as 35% with a bachelor’s degree or equivalent, 35% with a more advanced graduate degree, and 30% reporting lower educational levels.
A power analysis was conducted to estimate the sample size necessary to detect a small effect size (consistent with other research findings in this area) of .16 with an alpha of .05, based on the regression models described below, with a power of .80. The necessary sample size was estimated at 72.
Measures
Parent emotion representations
Parent emotion representations were measured using two questionnaires. The TMMS (Salovey et al., 1995) was used to measure parents’ perception, understanding, and acceptance of their own emotions. Parents used a five-point Likert scale to indicate how much they agreed with 30 statements about emotions. The clarity subscale was used to measure parents’ perception and understanding of their own emotions. Items in this subscale reflect the ability to recognize and comprehend one’s own mood (e.g., “I am usually very clear about my feelings”). The attention subscale was used to measure parental beliefs about the importance of attention to, and acceptance of, emotional experiences (e.g., “Feelings give direction to life”). The repair subscale provided a measure of parents’ effort to reduce their negative moods and maintain positive moods (e.g., “I try to think good thoughts no matter how badly I feel”). The TMMS is often used as a measure of emotional intelligence; prior studies using the TMMS have confirmed predicted associations between high subscale scores and independent measures of coping with stress, adaptive emotion management, and psychological well-being (Mayer & Salovey, 1995; Salovey, Stroud, Woolery, & Peel, 2002). Together, these questionnaires assessed how parents represent emotion in their own experience.
The ERQ (Gross & John, 2003) was used to measure parent emotion regulation style. Using a seven-point Likert scale, mothers were asked to rate how much they agreed with 10 statements that reflect two emotion regulation styles. The reappraisal style describes people who try to control their emotions by employing cognitive strategies (e.g., “When I want to feel more positive emotions, I change the way I’m thinking about the situation”). The suppression style describes people who try to control their emotions by inhibiting emotionally expressive behavior (e.g., “When I am feeling negative emotions, I make sure not to express them”). A large research literature using this measure indicates that there are significant correlates of these two approaches to emotion regulation with differences in interpersonal functioning and individual adjustment (John & Gross, 2004, 2007). Other work shows that the measure has good internal reliability, test–retest reliability, and measurement equivalence across gender and ethnicity (Gross & John, 2003; Melka, Lancaster, Bryant, & Rodriguez, 2011).
Parent expressivity
Parents self-reported on two dimensions of their socialization of emotion in the home. The first dimension captured indirect socialization through emotional expressivity within the family. It was measured with the Self-Expressiveness in the Family Questionnaire (SEFQ; Halberstadt, Cassidy, Stifter, Parke, & Fox, 1995). Parents rated the frequency that they express positive emotion in 20 common family situations (e.g., spontaneously hugging someone; expressing sympathy for someone’s troubles) and negative emotion in 20 typical situations (e.g., showing contempt for someone’s actions; crying after an unpleasant disagreement) on a nine-point Likert scale. A three-factor structure (positive, negative submissive, and negative dominant) was used for analyses, based on prior research. A number of studies find that positive and negative family emotional expressiveness is associated with children’s emotional expressiveness and emotion understanding in expected ways, and that the expression of negative dominant emotions (such as anger and hostility) is more difficult for children than the expression of negative submissive emotions (such as sadness and distress) because the former are more threatening (see Halberstadt, Crisp, & Eaton, 1999, and Halberstadt & Eaton, 2003, for reviews). Psychometric studies indicate that the measure is internally consistent and moderately stable over time (Halberstadt et al., 1995).
The second dimension captured more direct parental socialization of emotion in the home through the ways parents reacted to their children’s negative emotions. This was measured using the Coping with Children’s Negative Emotions Scale (CCNES; Eisenberg & Fabes, 1994; Fabes, Poulin, Eisenberg, & Madden-Derdich, 2002). Parents reported their reactions to the child’s expression of negative emotion for 12 hypothetical scenarios. The scenarios were designed around common situations in which children experience sadness, anger, fear, embarrassment, disappointment, and anxiety. On a seven-point Likert scale, parents rated the likelihood of reacting in six ways. These included: (1) matching the distress of the child (e.g., get angry with my child); (2) expressing punitive reactions that reduce the child’s exposure or need to address the negative emotions (e.g., send my child to his room to cool off); (3) encouraging the child to express her or his negative emotions or validating the expression (e.g., encourage my child to express her feelings of frustration); (4) emotion-focused interventions (e.g., soothe my child and do something fun to make him feel better); (5) focus on helping the child engage in problem-solving related to the causes of the emotion (e.g., help my child think of ways she can still be with friends); and (6) minimizing the seriousness of the event, the child’s problem, or reaction (e.g., tell my child to not make a big deal out of it). The CCNES is internally reliable and shows good test–retest reliability, and has been found to predict children’s emotional expressiveness within the family and social competence with peers (Fabes et al., 2002; McElwain, Halberstadt, & Volling, 2007)
Children’s emotion regulation
Children’s use of emotion regulation strategies was assessed using the Children’s Emotion Regulation Processes Survey (CERP; adapted from Bernzweig, Eisenberg, & Fabes, 1993, and Eisenberg et al., 1993). The original survey was adapted in two ways. Firstly, because this survey was originally designed for kindergarteners, strategies were adapted to be age-appropriate for slightly younger 4–5-year-olds (e.g., wording was simplified, and one category was deleted: “Actively tells himself/herself that it isn’t really something to get upset about” because it reflects more advanced self-regulatory capabilities). Secondly, the original survey was expanded from two hypothetical vignettes to eight to incorporate a broader range of self-regulatory challenges. Parents were presented with eight scenarios about everyday conflicts centered around the emotions of sadness and anger (e.g., “When your child is mad because another child took his/her toy and won’t share, how likely is your child to … ”) and a list of potential coping responses. Parents were asked to rate the likelihood that the target child would engage in each coping response on a seven-point scale. Coping responses were designed around three broad dimensions of the original questionnaire: (1) attention-focused (e.g., do something else to forget about it, like play a new game or talking about something else; think about something else to forget about it); (2) emotion venting (e.g., cry to release his/her frustrated feelings; hit or yell at the child so they give him/her the toy); and (3) constructive (e.g., approach the child or an adult to ask them why she can’t have the toy, and if s/he could have it later; ask another person for help or to intervene). Each dimension contained four response items. This measure was later factor analyzed to create internally consistent subscales as described below.
Data analysis
Descriptive statistics were calculated to assess distribution characteristics and concordance with mean and range values obtained from prior research using these measures. Intercorrelations were used to examine data reduction options as well as to evaluate bivariate associations among the study variables in preparation for subsequent analyses. Following preliminary analyses of differences in the study measures according to child gender and age, a series of multiple regressions were conducted to examine mediational relations between maternal emotion representations, the use of emotion-related socialization practices, and children’s emotion regulation strategies, consistent with the research hypotheses.
Results
Preliminary analyses
Table 1 presents descriptive statistics for all study measures, including means and standard deviations, the range of scores, internal consistency, and the number of items per measure. Univariate statistics and scale reliabilities for all parental emotion representation and socialization measures were examined and compared with previous studies. The ERQ and TMMS scores fell within the expected range. Furthermore, the intercorrelations between ERQ and TMMS variables replicated patterns found in past research (Gross & John, 2003). Attention and clarity were both negatively correlated with suppression (r = –.38; r = –.32, ps < .05), and clarity and repair were positively correlated with reappraisal (r = .35; r = .53, ps < .05). These findings suggest that parents’ pattern of responding was consistent across the different subtypes of parental representation dimensions.
Descriptive statistics for parental emotion representations, parental reactions to children’s negative emotions, and parental expressivity.
a Salvoey et al., 2002; b for female participants from Gross & John, 2003; c Fabes et al., 2002; d Valiente et al., 2004.
TMMS: Trait Meta-Mood Scale; ERQ: Emotion Regulation Questionnaire; CCNES: Coping with Children’s Negative Emotions Scale; SFEQ: Self-Expressiveness in the Family Questionnaire.
Descriptive statistics for the CCNES and the SEFQ also appear in Table 1. Replicating prior research by Eisenberg and Fabes (1994), the mean of scores for positive reactions (i.e., emotion-focused, problem-focused, and emotion encouraging) were higher than those of negative reactions (i.e., minimizing, punitive, and matched distress). Similarly, parents reported expressing more positive emotions than negative on the SEFQ, also replicating the pattern found in previous research (Halberstadt et al., 1995).
Data reduction
Due to the adaptation of the original form, the CERP items were factored using principal axis factor analytic techniques with varimax rotation. Four factors were retained under the criteria of having a factor loading higher than .3 and conceptual coherency (see Table 2): problem- and emotion-focused strategies (e.g., ask an adult for an alternative solution), attention-focused strategies (e.g., think about positive things), dominant venting strategies (e.g., hit or yell to obtain one’s goal), and submissive venting strategies (e.g., cry to release feelings). The internal consistency of each factor was in an acceptable range (Cronbach’s α = .70–.93).
Varimax rotated factor loadings and descriptive statistics for children’s emotion regulation processes questionnaire.
To reduce the number of variables for analyses, the correlations between the ERQ and the TMMS subscales were examined. The emotion regulation subscales, repair and reappraisal, were strongly correlated (r = .53, p < .001). Given their statistical association and conceptual similarity, these variables were rescaled and aggregated to form a composite “regulation” variable. Conceptually similar variables in the CCNES were also composited. Minimizing and punitive reactions were strongly correlated (r = .66, p < .001) and have been aggregated in past research (Eisenberg & Fabes, 1994). These variables were aggregated to form a “minimizing/punitive reaction” composite variable. Emotion-focused and problem-focused reactions were also aggregated based on a robust correlation (r = .53, p < .001) and similar associations with other variables to form a “problem-emotion focused reaction” composite variable.
Gender and age differences
Children’s gender and age were examined in relation to parental emotion representation variables, parent socialization variables, and children’s emotion regulation strategies. There were no gender differences and one age difference; parents reported that they expressed more positive emotions with younger children than with older (r = –.44, p < .05).
Bivariate correlations
Bivariate correlations were examined to understand the network of associations among the variables and to determine testable multivariate models (see Table 3). Besides significant associations among variables within each predictor set, consistent with prior research, there were also significant associations between measures of parent emotion representations, parent socialization practices, and child emotion regulation.
Bivariate correlates among parental emotion representation, socialization behaviors, and children’s emotion regulation strategies.
+ p < .10, * p < .05, ** p < .001.
TMMS: Trait Meta-Mood Scale; ERQ: Emotion Regulation Questionnaire; CCNES: Coping with Children’s Negative Emotions Scale; SFEQ: Self-Expressiveness in the Family Questionnaire.
Parental emotion representations were predictably associated with their reactions to children’s negative emotions. Dimensions of attention, clarity, and regulation were positively correlated with supportive and negatively correlated with unsupportive reactions. Specifically, parents higher in attention and regulation reacted in more problem- and emotion-focused and emotion encouraging ways, while parents scoring higher in clarity responded with more emotion encouragement. Parents who were higher in attention were also less minimizing/punitive, while those higher in clarity and regulation expressed less matching distress. Parents with higher scores on suppression were lower in emotion encouragement, but suppression scores were unrelated to any unsupportive socialization processes.
Parents higher in attention, clarity, and regulation also expressed more positive emotions in the home, while those higher in suppression expressed less. In addition, parents who had higher scores on regulation expressed fewer negative dominant emotions. These data on parental regulation suggest that parents who were focused on regulating their negative emotion in active ways did so for hostile negative emotions only.
With respect to children’s emotion regulation, there were fewer associations. Parents who valued attending to emotion had children using greater problem and emotion-focused strategies, and parents who managed their own feelings had children enlisting greater attention-focused strategies. Parents’ use of problem-emotion focused strategies was positively associated with children’s use of problem and emotion-focused, attention-focused, and submissive venting emotion regulation strategies. Parents’ emotion encouragement was also positively associated with children’s use of problem and emotion-focused strategies. Positive emotion expressed in the home was associated with children’s attention-focused self-regulatory strategies, and the expression of negative-submissive emotions in the home was associated with children’s problem and emotion-focused regulatory strategies.
Multivariate analyses
As earlier noted, two significant direct associations between parental emotion representations and children’s emotion regulation strategies were found. Parents higher in attention had children who were more likely to use problem and emotion-focused strategies. Parents higher in regulation were more likely to have children who engaged in attention-focused strategies. No other direct associations existed between parental emotion representations and children’s emotion regulation strategies.
Following Holmbeck’s (1997) criteria, if (a) parental emotion representations (predictors) were significantly associated with parental socialization behaviors (mediators), (b) parental emotion representations (predictors) were significantly associated with children’s emotion regulatory strategies (dependent variables), and (c) parental socialization behaviors (mediators) were significantly associated with children’s emotion regulatory strategies (dependent variables), then a mediation model qualified for testing. Mediation models were tested using multiple regression analyses, with parenting emotion representations and socialization behaviors as predictors, and children’s emotion regulatory processes as dependent variables, to determine if the impact of parent emotion representations is reduced after controlling for parent socializing behaviors. A series of four multiple regression models met Holmbeck’s criteria based on correlation data and were tested for mediation effects. In each model, three criteria were used to demonstrate a mediation effect when both predictors were entered into the model: (1) a significant contribution of the mediator to the model; (2) an increase in the amount of variance accounted for by the model with the addition of the mediator; and (3) a reduction in the contribution of the predictor to a non-significant value when the mediator is added. In addition to the four models, an aggregate model was tested based on correlations and conceptual coherency among variables.
Predicting children’s use of problem- and emotion-focused strategies
Two multiple regressions were conducted to test if parental reactions and expressivity mediated the relations between parental attention and children’s use of problem- and emotion-focused strategies. In each model, parental attention was entered in the first step and significantly predicted children’s problem and emotion-focused strategies (R 2 = .08, F(1, 71) = 5.67, p < .05). In the first model, the addition of parental problem-emotion-focused reactions to children’s negative emotions significantly contributed to the model, increasing the amount of variance accounted for in the model, and reduced the contribution of parental attention to a non-significant value (R 2 = .13, F(2, 70) = 5.02, p < .05); see Table 4(a)). In the second model, parental emotion encouragement had the same pattern of influence on the relation between parental attention and children’s problem and emotion-focused strategies; however, the significance of emotion encouragement and the ▵R 2 were both marginal (R 2 = .11, F (2, 71) = 4.49, p < .05; see Table 4(b)).
Summary of multiple regression analyses for mediated models predicting children’s emotion regulation strategies.
(a) Predicting children’s problem and emotion-focused strategies.
R 2 = .08 for Step 1; ▵R 2 = .05 for Step 2 (ps < .05).
* p < .05.
(b) Predicting children’s problem and emotion-focused strategies.
R 2 = .08 for Step 1; ▵R 2 = .04 for Step 2 (p < .08).
+ p < .08; * p < .05.
(c) Predicting children’s attention-focused strategies.
R 2 = .07 for Step 1; ▵R 2 = .11* for Step 2 (p < .08).
* p < .05.
(d) Predicting children’s attention-focused strategies.
R 2 = .07 for Step 1; ▵R 2 = .05 for Step 2 (ps < .05).
* p < .05.
(e) Predicting children’s constructive emotion regulation composite.
R 2 = .09 for Step 1; ▵R 2 = .08 for Step 2 (ps < .05).
* p < .05.
Predicting children’s use of attention-focused strategies
A series of two multiple regressions tested the impact of socialization behaviors on the relation between parental emotion regulation and children’s use of attention-focused strategies. In each model, parental regulation was entered in the first step and significantly predicted children’s attention-focused regulation strategies (R 2 = .07, F(1, 71) = 5.04, p < .05). In both models, parental problem-emotion-focused reactions (R 2 = .18, F(2, 70) = 7.55, p < .05; see Table 4(c)) and positive expressivity (R 2 = .12, F(2, 70) = 4.72, p < .05; see Table 4(d)) significantly increased the amount of variance accounted for in each model, significantly contributed to the model, and reduced parental emotion regulation to a non-significant contributor.
Aggregate model
To test a more generalizable model of emotion socialization that predicted children’s constructive approaches to emotion regulation, variables were further aggregated based on the correlations among variables, the individual pathways tested in the four multiple regressions described above, and conceptual coherency. The aggregate parental “emotionally supportive representations” was constructed from parental attention and regulation variables (r = .22, p < .05). Higher scores on this aggregate represent parents who are attentive to their own emotional states, believe emotions are important in their lives, and put forth effort to repair negative moods, maintain positive emotional states, and use active approaches to emotion regulation (i.e., reappraisal).
Parental “positive and supportive socialization behaviors” was created by rescaling and compositing parental positive expressivity, emotion encouragement, and problem and emotion-focused reaction variables. Parents with higher scores on this composite reflect those who exhibit a positive emotional environment and promote an active approach to helping their child navigate negative emotional states. Positive expressivity was moderately correlated with both emotion encouragement (r = .41, p < .001) and problem- and emotion-focused reactions (r = .46, p < .001). Emotion encouragement and problem- and emotion-focused reactions were also moderately associated (r = .44, p < .001).
“Children’s constructive emotion regulation strategies” was constructed from aggregating problem- and emotion-focused strategies and attention-focused strategies from parent report. These variables exhibit strong conceptual coherency, as well as similar patterns of association across other variables and were therefore aggregated to represent a single style of emotion regulation strategy, although they were not statistically related. Together, these strategies represent children who actively put forth an effort to reduce their negative affect by employing resources within the self or by eliciting social support.
The three composite variables were significantly associated. Parents’ emotional supportive representations were associated with parents’ positive and supportive socialization behaviors (r = .64, p < .001) and children’s constructive emotion regulation strategies (r = .30, p < .05), and parents’ positive and supportive socialization behaviors also correlated significantly with children’s constructive emotion regulation strategies (r = .41, p < .05).
The purpose of the aggregate model was to test the mediating impact of positive and supportive socialization behaviors on the relation between parents’ emotionally supportive representations on children’s constructive emotion regulation strategies. These new composite variables met the criteria for mediation testing as stated by Holmbeck (1997). In the first step, parental emotionally supportive representations significantly predicted children’s constructive regulation strategies (R 2 = .09, F(1, 71) = 6.89, p < .05). When parental positive and supportive socialization behaviors were added to the model in the second step, they contributed significantly to the model, increased the amount of variance accounted for in the model, and reduced the contribution of emotionally supportive representations to a non-significant value (R 2 = .17, F(2, 70) = 7.25, p < .001; see Table 4(e)). Parental positive and supportive socialization behaviors mediated the association between parents’ emotionally supportive representations and children’s constructive emotion regulation strategies.
Discussion
The results of this investigation identify two kinds of parental emotion representations that are important to their socialization practices and children’s emotion self-regulation. Parents’ values about the importance of attending to and accepting emotional experiences, and their beliefs in regulating negative moods and maintaining positive emotion, were each associated with more supportive emotion socialization efforts and with children’s constructive self-regulatory strategies (see Yap et al., 2008, for similar results with a preadolescent sample). Furthermore, the associations of these parental emotion representations with children’s self-regulation were mediated by these socialization processes, consistent with theoretical expectations, such as the concept of the parent meta-emotion philosophy of Gottman et al. (1996, 1997).
The importance of these representational processes is noteworthy. First, parents who believe that emotions merit attention and consideration, and accept them as valid indications of personal well-being, would be more likely to devote comparable attention to the feelings of their children and to regard them as legitimate rather than dismissing or minimizing their importance. Perhaps for this reason, parents who represented emotions in this manner were also more likely to encourage their children’s emotional expressions, engage them in problem-solving or emotion-focused management, maintain an emotionally positive family environment, and were less likely to respond to children’s negative emotions with punitive or minimizing responses. Wong and colleagues (2008) also noted that parents who believed in the importance of children’s feelings were more likely to encourage their expression.
Secondly, parents who value emotion self-regulation, often through reappraisal, would be more likely to actively assist in their children’s emotional self-management, and to become models for their children of emotional self-efficacy (Saarni, 1999). Consistent with this view, parents who represented emotions in this manner were also more likely to engage children in emotion-related problem-solving, encourage their emotional expression, foster a positive family environment, and were less likely to match children’s negative emotions with their own negative reactions or enlist negative dominant emotions (such as anger) in the family environment. Both attention and regulation beliefs were associated also with children’s problem- and emotion-focused and attention-focused self-regulation strategies. A third aspect of parents’ emotion representations, consisting of beliefs in the importance of clarity in recognizing and comprehending one’s feelings, had comparable associations with constructive emotion socialization processes, but clarity was not associated with children’s emotion self-regulation.
A fourth aspect of parent emotion representations assessed in this study – endorsing the value of suppressing or inhibiting emotionally expressive behavior – did not have significant associations with unsupportive emotion socialization processes (such as minimizing or punitive reactions) nor with an emotionally negative family climate. Instead, these beliefs were more distinctly associated with what parents were not doing. Parents higher in emotional suppression were less likely to encourage their children to express their emotions, consistent with their own expressive style, and they were less likely to maintain a positive emotional climate in the family. This is important for underscoring that certain parent emotion representations can be influential by undermining constructive emotional socialization influences, as well as by fostering negative ones. Considerable research underscores the importance, for example, of positive emotional expressiveness in the family in supporting children’s emotional competence (e.g., Halberstadt & Eaton, 2003; Valiente, Fabes, Eisenberg, & Spinrad, 2004). In this study, parental suppression of emotion was not associated with negative emotion socialization practices, but rather with a reduction in positive influences. Parents endorsing unsupportive emotion socialization practices or reporting an emotionally negative family environment were, in turn, distinctive by their lack of attention and acceptance of their own feelings, emotion self-regulation, and clarity in understanding their feelings and their significance.
The importance of parental emotion representations emphasizing attention to and regulation of emotional experience is underscored by the mediational models associating these representations with parental socialization practices and children’s emotion self-regulation. Consistent with theoretical expectations, these models indicated that parental problem- and emotion-focused socialization strategies significantly mediated (and parental encouragement of emotional expression marginally mediated) the association between parental attention and children’s emotion- and problem-focused self-regulatory strategies. Although causal conclusions cannot be derived, and the direct association between parental emotion representations and children’s emotion regulation strategies was not strong, these findings are consistent with the view that parental attention to and acceptance of children’s emotions motivates efforts to affirm their expression and engage children in problem-solving, with the latter possibly functioning as a model for children’s own problem-focused emotion regulatory strategies. In a similar manner, parental problem/emotion-focused socialization strategies and positive expressivity in the family significantly mediated the association between parental regulation and children’s use of attention-focused self-regulatory strategies. These findings are consistent with the view that parents’ beliefs in the importance of emotion management motivates their engagement with children in emotion problem-solving strategies (and helps to maintain a positive family environment), which may also guide children’s self-regulatory efforts focused on attentional diversion. Taken together, this is the first study to test and confirm theoretical models of the association of specific parent emotion representations with children’s emotion self-regulatory strategies mediated by parents’ emotion-related socialization practices.
The final, aggregate model provided support for the more general view that parents’ emotionally supportive representations are associated with children’s constructive emotion regulation strategies through their influence on parents’ positive and emotionally supportive socialization behaviors. The aggregate model and the associations among the variables composing each composite are important for underscoring that the components of adult emotion representations and emotion-related socialization practices assessed in this study are not independent but important as constellations of emotion-related influences within the family.
Other findings from this study merit note for what they reveal about the family emotional climate affecting children’s emotion regulation. Parental positive and negative submissive expressivity were positively correlated, and the latter also predicted children’s problem- and emotion-focused self-regulation. Similarly, children’s submissive venting was predicted by parents’ problem-emotion-focused reactions to children’s emotions. Notably, none of these relations was observed for the expression of dominant negative emotions in the family, supporting the conclusion that submissive negative emotional expressions potentially have more psychologically constructive consequences than the expression of dominant negative emotions because the latter elicit perceptions of threat and danger (Thompson & Meyer, 2007). This merits further exploration as a way of better understanding the diverse expressions of negative emotion in the family and their influences on children’s self-regulation.
In sum, these findings provide the basis for further examination of specific components of parental meta-emotion philosophies and their associations with emotion-related socialization practices. It is important to note that these findings do not support the view that young children simply appropriate parents’ emotion self-regulatory style as their own, but rather that their developing capacities for emotion self-regulation emerge from multiple features of parental responses to their feelings. The effort of this investigation to assess several central components of parent emotion representations – attention/acceptance, clarity, regulation, and suppression – highlights the importance of studying other aspects of emotion-related representations that may also be relevant to emotion socialization in the home, including the evaluation of intense emotion (e.g., how one feels about becoming emotionally aroused) and beliefs about specific negative emotions (such as anger and sadness) for which gender differences are likely to be apparent (Fivush, 1998). Although there were very few age differences in this sample of narrow range, it would not be surprising if different aspects of a parental meta-emotion philosophy would be particularly relevant for children of different ages as the child’s self-regulatory capabilities, the parent’s socializing role, and the young person’s capacities for autonomy change with age. In sum, further understanding of parental values and beliefs concerning emotion, and their association with children’s emotion socialization, is warranted in efforts to better model emotional growth in childhood.
Limitations
The interpretation of these findings is limited in several ways. Firstly, the sample reflects the socioeconomic and racial diversity of the recruitment area but is not more broadly representative (particularly as this was a highly educated sample), and thus the findings may not be fully generalizable. Secondly, the regression models evaluated in this research were theoretically based but relied on measures gathered contemporaneously, and consequently causal conclusions are not warranted. Although it is reasonable to expect that parents’ representations guide their behavior toward children which, in turn, influence children’s self-regulation, and such a view is supported by considerable research in family socialization (see, e.g., Cummingham et al., 2009), alternative models are reasonable and should also be examined in future research. For example, although parents’ beliefs about emotion may guide their responses to children’s negative emotions, it is also reasonable that parental beliefs emerge out of parents’ past encounters with their children in emotion-laden encounters, and thus the causal direction may be different from the one identified in this study. Moreover, it is also possible that parents’ beliefs and socialization processes related to children’s emotions may vary for different emotions: the socialization of anger and its expression may be different from socialization processes associated with sadness, for example.
Thirdly, the data from this study are based on parental report, and thus these findings are subject to the possibility of shared source variance and shared method variance. The measures used in this study were selected for their validational evidence and strong psychometric properties, and descriptive statistics and correlational findings from this study are consistent with other research using such methods that have also relied on parental report (e.g., Wong et al., 2009). Nevertheless, multi-method studies inspire greater confidence because they avoid the potential problems of informant bias and shared method variance. In particular, research predicting children’s self-regulation of emotion should enlist direct observations of children’s emotion regulation to confirm the findings reported here.
Despite these limitations, this study is among the first to identify specific features of how parents represent emotion in their own lives and their relevance to how they socialize emotion in children. It contributes further understanding to the meta-emotion philosophies of parents and the processes by which emotion and emotion regulation are developmentally guided by family experience. These findings also suggest useful new directions for study in the socialization of emotion in children.
Footnotes
Funding
The research reported in this paper was supported by a generous grant from The Amini Foundation for the Study of Affects.
