Associations Between Emotional Competence and Prosocial Behaviors With Peers Among Children With Autism Spectrum Disorder

Abstract

Children with autism spectrum disorder (ASD) are at risk for disrupted peer interactions. This study contributes to our understanding of how multiple foundational elements of emotional competence are related to children’s prosocial behaviors with peers. Children with ASD demonstrated significantly lower non-stereotypical affective perspective taking, had lower ratings of emotion regulation, and showed differences from their typical peers in the use of discrete coping strategies during peer interactions. Children’s emotion regulation and use of discrete coping strategies in the context of peers were associated with their prosocial behaviors one year later. The findings add to our understanding of how emotional development contributes to individual differences in the social-emotional behaviors of children with ASD. Implications for intervention are discussed.

During the preschool years, peer interactions can be emotionally charged situations. Learning to respond in socially appropriate ways to these situations is a key developmental task for preschoolers. Social competence is reflected through such behaviors as cooperation, showing kindness, and offering help to others (Ladd & Profilet, 1996). Socially competent children tend to have more supportive relationships with peers (Ladd et al., 1999), which are an important source of assistance and security throughout the school day (Hartrup, 1996). Unfortunately, for children with autism spectrum disorder (ASD), social interactions are an area of significant challenge. Even children with more advanced cognitive and language skills show difficulties with peer engagement, including for example, less frequent and poorer-quality social initiations and responses with peers and more ritualistic behaviors. As a result, they may have more disrupted exchanges, fewer reciprocal friendships, and hold less-central positions in peer networks (Bauminger, 2002; Chamberlain et al., 2007). There is, however, important heterogeneity in initiations of social exchanges with peers among children with ASD. Many children with ASD are inclined towards social interactions and friendships but have difficulty engaging successfully (Bauminger et al., 2003), pointing to the need to understand factors linked to individual differences in their social competence with peers. Scholars have proposed a model identifying the emotional prerequisites for typical preschoolers’ social competence with peers, that is, those indices of emotional competence that have lasting effects on children’s social outcomes (Denham et al., 2003; Rose-Krasnor, 1997). These prerequisites include children’s knowledge or understanding about emotions (Denham & Couchoud, 1990a), emotion regulation or the modulation of emotional reactivity (Calkins & Mackler, 2011), use of effective coping strategies in the context of peer interactions (Eisenberg et al., 1993), and emotional expressiveness (e.g., temperamental negative affect; Rothbart, 2006). In line with Denham et al.’s (2003) notion that these emotional prerequisites should be studied in a comprehensive rather than piecemeal approach, the present study examined group differences in children with and without ASD in each of these indices of emotional competence, and explored patterns of association between children’s emotional competence and prosocial behaviors with peers one year later.

Emotion Knowledge

One of the key capacities associated with children’s emotional competence is their knowledge about their own and others’ affective experiences. The ability to discriminate and label simple emotions such as happiness, sadness, anger, and fear, emerges by around age two for typically developing children (Denham & Couchoud, 1990a), whereas more complex emotions like pride, embarrassment, guilt, and empathy emerge later, once children begin to show self-evaluation, social comparison, and perspective-taking (Denham, 2007). By the end of preschool, children appreciate the association between emotion expressions, situations that cause emotions, and consequences of such emotional expressions (Denham & Couchoud, 1990a), and by kindergarten children realize that there are equivocal situations in which other children may have different emotional reactions than themselves (Denham, Caverly, et al., 2002; Denham & Couchoud, 1990b). Importantly, the development of emotion recognition, especially that regarding complex emotions, appears to coincide with children’s increased prosocial behaviors in social contexts; for example, from around three to five years, most children are able to respond with positive interventions when others are in need (Denham, McKinley et al., 1990) and emotional knowledge predicts social competence, both concurrently and in kindergarten (Denham et al., 2003).

In contrast to the typical course of development, some aspects of emotion knowledge appear to be delayed for young children with ASD. Although findings on group differences have been mixed (e.g., Humphreys, 2007; Ozonoff, 1990), there is evidence that children with ASD have specific delays in attending to relevant emotional facial cues in others, as well as recognizing and interpreting the emotional expressions of others due to deficits in processing emotion-related stimuli (e.g., Hobson et al., 1988; Miranda et al., 2017). Yet, there is also evidence of relatively less difficulty with basic understanding of those emotions (e.g., Downs & Smith, 2004), relative strengths with respect to recognizing simple emotions (e.g., Heery et al., 2003), and relatively more challenges with respect to negative, ambiguous, complex, and social emotions (e.g., Bal et al., 2010; Capps et al., 1992; Heery, et al., 2003; Williams & Happé, 2010). Among studies that reported group differences, children with ASD have shown poorer performance on emotion recognition for negative emotions like “angry” and “afraid/fear” in contrast to “happy” (Lozier et al., 2014; Nuske et al., 2013; Rump et al., 2009). Furthermore, challenges associated with knowledge of specific emotions appear to have particular consequences for the social competence of children and adults with ASD. For example, recognition of sadness was associated with socialization scores on the Vineland Adaptive Behavior Scales II for 4 to 7-year old children with ASD (VABS II; Sparrow, 2005; Williams & Gray, 2012), to parents’ reports of adaptive function among adolescents with ASD (Wallace et al., 2011), and to social competence, as measured by the reciprocal social interaction subscale of the Autism Diagnostic Observation Schedule for adults with ASD (Boraston et al., 2007). Together, these findings point to the need to better understand how individuals with ASD apply knowledge of emotions to social situations. In the present study, we examined these associations in preschoolers with ASD.

Emotion Regulation and Coping Strategies

Emotion regulation refers to the ability to modulate emotional experiences or expressions to achieve individual goals or social adaptation (Calkins & Hill, 2007; Eisenberg & Spinrad, 2004), and is believed to be a function of emotional reactions and the ability to deploy effective strategies to cope with such reactions (Cole et al., 2004). Challenges with emotion regulation predict poor interpersonal interactions and reduced social competence in typically developing preschoolers (Calkins & Mackler, 2011). Typical preschoolers’ discrete coping strategies in emotionally-charged social situations, which include distracting themselves from a negative situation, avoidance, making their dissatisfaction known with aggressive behaviors or venting (e.g., crying), seeking support from others, and constructive or reframing techniques like instrumental coping and cognitive restructuring, are critical for maintaining positive exchanges and forming enduring relationships with peers (Eisenberg & Spinrad, 2004). Mothers’ reports of children’s typical coping strategies are related to their social competence in the school setting (Eisenberg, et al., 1993; Fabes et al., 1994).

Emerging evidence suggests that emotion regulation difficulties are linked to many of the behavioral challenges associated with ASD (Jahromi et al., 2013; Laurent & Rubin, 2004; Loveland, 2005; Prizant et al., 2003; Samson et al., 2015; Wetherby et al., 2000). Parent and teacher reports point to self-regulatory difficulties for children with ASD, even among children with no intellectual disability (Gomez and Baird, 2005; Ashburner et al., 2010), and recent population-based work seems to support the notion that the majority of children with ASD show emotional problems (Totsika et al., 2011).

With regard to discrete coping strategies employed by young children to regulate their emotions, children with ASD have been shown to be less likely to use advanced strategies (e.g., complying) and more likely to use less advanced strategies (e.g., defending, objections, crying/venting) in the context of emotional arousal (Konstantareas & Stewart, 2006). When faced with frustration in particular, preschoolers with ASD showed significantly more avoidance and venting strategies, and significant fewer constructive strategies compared to their typical peers (Jahromi et al., 2012). Finally, adolescents with ASD show no differences from their typical peers in self-reported adaptive emotion regulation strategies (e.g., problem solving) but they report significantly more maladaptive strategies (e.g., rumination, “shutting down”), thus it may be that children with ASD have continued challenges avoiding the more maladaptive coping strategies to regulate emotion (Mazefsky et al., 2014). In the present study, we extended this work to the peer context by measuring both children’s emotion regulation and the coping strategies they employ in the context of peer interactions to assess relations between these aspects of emotional competence and children’s prosocial peer interactions.

Temperamental Emotional Expressiveness

Children’s emotional expressiveness is the final index of emotional competence identified by Denham et al. (2003) as a prerequisite of children’s social competence. According to Rothbart and colleagues’ conceptualization of temperament, negative affect refers to the frequency, intensity, and duration of negative experiences and expressions (Rothbart et al., 2006). In the context of emotionally charged peer interactions, children with less temperamental negative reactivity may be better at expressing less intense reactions and maintaining more positive peer interactions (e.g., Goldsmith et al., 2001; Rudasill et al., 2013). Findings on temperamental ratings of children with ASD have been mixed (see Jahromi, 2017). Parent reports suggest that children with ASD are rated as having more negativity and “difficult” behaviors, including more negative affective expressions like sadness, fear, anger, shame, and guilt (Bailey et al., 2000; Capps et al., 1992; Kasari & Sigman, 1997; Hepburn & Stone, 2006; Kasari et al., 2012; Kasari et al., 1990), difficulties in sub-factors associated with negative affect, like sooth-ability and perceptual sensitivity (Samyn et al., 2011, for a review), and in measures of adaptability and mood (Chuang et al., 2012). On the other hand, others have found no differences in the negativity ratings of children with and without ASD at younger ages (Konstantareas & Stewart, 2006). The discrepant findings warrant further investigation as they may be related to the age and level of functioning of study samples; differences may be less apparent at younger ages, but may become more pronounced as children with ASD fail to keep pace, developmentally, with increasing social-emotional challenges.

There may also be important social consequences of the negative affect that is often reported among children with ASD. Scholars argue that the emotional arousal experienced by children with heightened negativity may undermine children’s ability to take in meaningful information from the social setting and to identify effective social problem solving strategies, which could ultimately lead to more aggression and peer rejection (Dollar et al., 2017; Pope & Bierman, 1999) and there is evidence of such a process in the study of typically developing preschoolers and kindergarteners (Denham et al., 2003). Thus, we aimed to extend this line of work by examining the association between temperamental negative affect and prosocial behaviors with peers among preschoolers with ASD.

The Current Study

There is a need to better understand factors related to the heterogeneity in social-emotional outcomes among young children ASD. Drawing from Denham et al.’s (2003) framework on the emotional prerequisites of young children’s social competence, the first goal of the study was to examine group differences in emotion knowledge, emotion regulation, coping strategies in the context of peers, and emotional expressiveness (i.e., negative temperament) among preschoolers with and without ASD to understand differences in these emotional prerequisites between groups. Second, we aimed to explore associations between emotional prerequisites and prosocial behaviors with peers among preschoolers with ASD to understand whether such associations emerged for this group.

Method

Participants

Participants included 40 children (36 males) with a mean age of 54.57 months at recruitment (SD = 11.31 months) and their parents. Participants were drawn from a university-based preschool and community resource center for ASD in a metro-politan area in the southwest. Participants with ASD were recruited via electronic recruitment materials distributed to parents of children with ASD at the resource center and to parents of typically developing children at the university preschools. The sample consisted of 20 children with ASD (M = 58.95, SD = 11.50 months) whose clinical diagnosis was confirmed with the Autism Diagnostic Interview (ADI-R; Lord et al., 1994), and 20 typically developing children (M = 50.20, SD = 11.12 months) who were matched on gender and expressive language as per Charman (2004). Inclusion criteria required children to have verbal language, including the ability to put together complex sentences and a receptive language age of at least 3 years. Children included in the study did not have comorbid diagnoses that would impact the data obtained in the study. All the children in the typical sample had scores below 15 (i.e., the cutoff criteria for possible ASD) on the Social Communication Questionnaire (SCQ), a companion screener with properties similar to the ADI-R (Rutter, 2003). There were no significant differences between the matched groups on mental age or receptive or expressive language (see Table 1 for descriptive data and group difference data on developmental variables). Participants were predominately White (77.5%), whereas 10% were Latino, 7.5% were Asian, and 5% identified their ethnicity as “other”. The majority of the parents (97.5%) were married. Mothers had a mean age of 36.3 years (SD = 4.92), and fathers had a mean age of 37.6 years (SD = 6.08). With respect to their highest level of education, 5% of mothers and 5% of fathers earned a high school diploma, 50% of mothers and 47.5% of fathers earned a college degree, and 45% of mothers and 47.5% of fathers earned a graduate degree.

Table 1.

Descriptive Statistics for Developmental and Study Variables

	Autism			Typical
	M	SD	Range	M	SD	Range	Group Differences
Developmental Level
Chronological Age	58.95	11.50	[40–77]	50.20	11.12	[33–78]	F (1, 38) = 5.98, p = .02, d = .77
Mental Age	57.78	16.80	[32–93]	52.95	13.66	[28–85]	F(1, 38) = 1.99, p = .32, d = .32
Expressive Language Age	56.70	12.36	[32–83]	58.05	12.01	[37–81]	F (1, 38) = .12, p = .72, d = .11
Receptive Language Age	60.20	13.53	[39–81]	58.05	11.63	[45–81]	F(1, 38) = .29, p = .59, d = .17
Emotion Knowledge
Affective Labeling	1.76	.43	[.25–2.0]	1.75	.41	[.75–2.0]	F (1, 38) = .002, p = .96, d = .02
Expressive Labeling	1.66	.50	[.25–2.0]	1.63	.55	[.00–2.0]	F (1, 38) = .02, p = .88, d = .05
Receptive Labeling	1.84	.48	[.00–2.0]	1.86	.39	[.50–2.0]	F (1, 38) = .02, p = .88, d = .04
Happy	1.90	.31	[1.0–2.0]	1.95	.22	[1.0–2.0]	F (1, 38) = .35, p = .56, d = .18
Sad	1.83	.49	[.00–2.0]	1.85	.36	[1.0–2.0]	F (1, 38) = .03, p = .85, d = .06
Angry	1.78	.47	[.00–2.0]	1.73	.63	[.00–2.0]	F (1, 38) = .08, p = .78, d = .08
Scared	1.52	.62	[.00–2.0]	1.47	.66	[.00–2.0]	F (1, 38) = .06, p = .81, d = .07
Affective Perspective Taking
Stereotypical	1.35	.67	[.00–2.0]	1.60	.255	[1.0–2.0]	F (1, 38) = 2.41, p = .13, d = .52
Non-Stereotypical**	1.12	.72	[.00–2.0]	1.66	.44	[.40–2.0]	F (1, 38) = 7.83, p = .01, d = .91**
Emotion Regulation**	66.45	8.30	[53–79]	73.80	6.91	[57–81]	F (1, 38) = 9.26, p = .004, d = .96**
Emotion Regulation	24.3	3.53	[19–30]	27.3	3.03	[21–31]	F (1, 38) = 8.33, p = .006, d = .93**
Lability/Negativity	32.6	7.04	[23–45]	27.5	5.49	[19–37]	F (1, 38) = 6.52, p = .015, d = .83**
Coping Strategies with Peers	3.30	.88	[1.6–4.7]	4.37	.62	[3.4–5.6]	F (1, 38) = 21.33, p <.001, d = 1.41***
Avoidance/Distraction	3.25	1.22	[1.8–6.2]	3.11	.87	[1.9–4.9]	F (1, 38) = .18, p = .67, d= .13
Aggressive Coping*	3.57	1.74	[1.0–6.0]	2.65	1.06	[1.0–4.6]	F (1, 38) = 4.07, p = .05, d = .64*
Venting	3.69	1.44	[1.0–6.3]	3.49	1.29	[1.0–5.6]	F (1, 38) = .22, p = .65, d= .15
Seeking Support**	3.49	1.70	[1.0–7.0]	4.72	1.12	[2.7–7.0]	F (1, 38) = 7.32, p = .01, d = 1.02**
Cognitive Restructuring**	1.37	.67	[1.0–3.0]	2.30	.98	[1.0–4.5]	F( 1, 38) = 12.21, p = .001, d = 1.11**
Instrumental Coping***	2.17	.89	[1.0–4.0]	3.56	1.18	[1.0–6.0]	F (1, 38) = 17.63, p < .001, d = 1.33**
Temperamental Negativity	2.43	.79	[.93–4.0]	2.57	.96	[.60–4.4]	F(1, 38) = .26, p = .61, d = .17
Prosocial Behaviors with Peers*	15.33	2.89	[11–21]	17.63	2.98	[11–21]	F (1, 38) = 5.65, p = .02, d = .78*

^*p < .05,

^**p < .01,

^***p < .001.

Procedure

Children and their parents attended two laboratory visits that occurred approximately one month apart (visits 1 and 2), and parents completed a follow-up questionnaire approximately one year later. At visit 1, we assessed children’s cognitive and language level as well as their emotion knowledge, while parents completed an ADI-R interview and filled out questionnaires. At visit 2, children and parents engaged in laboratory activities and parents completed additional questionnaires. The majority (95%) of parents completed the follow-up one year later (Mage = 50 months; SD = 10.08 months).

Measures

Developmental and Diagnostic Status

Language age was measured using the Preschool Language Scale 4 (PLS-4; Zimmerman et al., 2002), from which we derived children’s expressive language age and receptive language age. This measure is designed for children from birth to 6.11 years and has been validated for use with children with autism (Zimmerman et al., 2002). Mental age was measured using the Differential Abilities Scale II (DAS-II; Elliot, 2007) which yielded a global composite ability (GCA) score from which we derived children’s mental age (mental age = [chronological age × GCA]/100). This measure is designed for children 2 years, 6 months to 17 years, 11 months, and has been validated for clinical samples with developmental delay (Elliot, 2007). To confirm the diagnosis of children with ASD, their parents completed the Autism Diagnostic Interview-Revised (ADI-R; Lord et al., 1994), a standardized, structured parent interview which assesses the presence and severity of symptoms of autism in early childhood. To confirm the status of children in the typical sample, their parents completed the Social Communication Questionnaire (SCQ; Rutter, et al., 2003), an ASD screener with properties consistent with the ADI-R; all children in the typical sample had scores below the cutoff criteria for possible ASD on the SCQ.

Emotion Knowledge

The Affective Knowledge Puppet Task (Denham, 1986) was used to assess children’s knowledge of emotion labels and affective perspective taking. To measure affective labeling, children examined four faces made of felt, which included expressions of happy, sad, angry, and afraid, and were asked to identify the appropriate emotion expressively (i.e., verbally label) as well as receptively (i.e., point to). Based on past work that highlighted the importance of children’s understanding of specific (i.e., negative) emotions for their social competence, we examined these emotion labeling variables separately.

To measure affective perspective taking, the puppets were used to enact 14 vignettes in which vocal and visual affective cues were paired with the puppets. Affective perspective taking items were broken into stereotypical affective perspective taking such that, during eight of the vignettes, the puppets expressed the typical emotion people would feel (e.g., fear during a nightmare). To measure non-stereotypical affective perspective taking, during six of the vignettes, the puppet was described as feeling the opposite of what the child would typically feel, based on mothers’ prior reporting in a questionnaire. After watching the vignette, the participant was asked, “How does the puppet feel?” and were told to place the correct face on the puppet.

Past research has supported the convergent validity of the measure by demonstrating associations between children’s performance on this task and other measures of emotion knowledge (Morgan et al., 2010), self-regulation (Denham et al., 2012), and children’s emotional expressiveness and social competence (Denham et al, 2003). Moreover, the task has successfully been used in prior research on children with developmental delays, including those with ASD (Travis et al., 2001), Down Syndrome (Kasari et al., 2003), and communication disorders (McCabe & Meller, 2004).

Emotion Regulation

Emotion regulation was assessed using the Emotion Regulation Checklist (ER Checklist; Shields & Cicchetti, 1997), a 24-item parent-report that yields two subscales: emotion regulation (e.g., “can modulate excitement in emotionally arousing situations”) and lability/negativity (e.g., “exhibits wide mood swings). Using a 4-point Likert-type scale (Rarely to Almost always), parents reported how often their child exhibits certain behaviors. The measure shows positive associations with children’s school and social adjustment (e.g., Shields et al., 2001) and has also been used in past work with children with ASD and has shown convergent validity with other measures of self-regulation and coping (Jahromi et al., 2013; Scarpa & Reyes, 2011). Emotion regulation and negativity/lability were examined separately. In addition, because we conceptualized emotion regulation to reflect greater control over emotional expressions, we also created an emotion regulation composite measure of the two subscales. The negativity/lability subscale was first reverse coded to reflect decreasing dysregulation before averaging the standardized subscale scores. The resulting variable reflected higher scores on the emotion regulation subscale and lower scores on negativity/lability subscale (Cronbach’s alpha for the present study = .84).

Coping Strategies in Emotionally Arousing Peer Interactions

The Parents Reports of Children’s Reactions (Eisenberg et al., 1994; Fabes et al., 1994) was used to assess the child’s responses to emotion-eliciting situations. This measure consisted of 44 items asking parents to report on their child’s typical coping strategies in response to specific situations that may require emotion regulation, (e.g., when the child has been ridiculed by a peer, when other children are playing without the child, and when a peer has disrupted the child’s activity). In the present study, we obtained adequate internal reliability on all subscales (current study’s Cronbach’s alphas are below). Responses reflected typical coping strategies employed by preschoolers: (1) seeking support (e.g., seeks the emotional support of an adult but not by crying; Cronbach’s alpha = .90), (2) cognitive restructuring (e.g., tells himself that it isn’t something to get upset about; Cronbach’s alpha = .87), (3) instrumental coping (e.g., asks the other children why they won’t let him play; Cronbach’s alpha = .80), (4) aggressive coping (e.g., physically or verbally tries to hurt others to release frustration; Cronbach’s alpha = .92), (5) distraction/avoidance (e.g., does something else to forget the problem; leaves the scene; Cronbach’s alpha = .85), (6) venting (e.g., cries to release frustration or to get adult intervention; Cronbach’s alpha = .90).

Convergent validity of this measure is supported by associations between ratings of children’s coping strategies and children’s heart rate variability, a physiological index of their emotion regulation (Fabes et al., 1994), observations of children’s social skills, and measures of children’s sociometric status among peers (Eisenberg et al., 1993). To our knowledge, the measure has not been used with children with ASD. In addition to examining the coping strategies separately, consistent with the authors of the measure, we also created a composite measure of coping strategies (i.e., an average of standardized scores) to reflect greater use of strategies that have been shown to be positively associated with social wellbeing (i.e., instrumental coping) and less use of strategies that were identified as negatively associated with children’s social wellbeing (i.e., acting out/aggressive strategies) in past work with this measure (Eisenberg et al., 1993).

Temperamental Negativity

Temperamental negativity was assessed using the Child Behavior Questionnaire–Short Form (CBQ-SF; Putnam & Rothbart, 2006; Rothbart et al., 2001), a parent-report temperament measure. Using a 7-point Likert-type scale (extremely untrue of my child to extremely true of my child), parents reported how true particular statements were about their child’s behaviors (e.g., “Has temper tantrums when s/he doesn’t get what s/he wants”). The measure has been widely used to assess parents’ reports of their children’s temperament, and a growing number of studies have demonstrated good internal reliability and convergent validity with their use of this measure for children with ASD (e.g., Adamek et al., 2011). Based on the factor structure proposed by Putnam and Rothbart (2006), the CBQ-SF negativity factor included fear, sadness, anger/frustration, distress, and low soothability (Cronbach’s alpha for the present study = .79).

Children’s Prosocial Behaviors With Peers

Social competence was assessed using the prosocial behaviors subscale of the Child Behavior Scale (CBS; Ladd & Profilet, 1996), which reflects the degree to which children were rated as engaging with their peers in a positive, socially-competent manner (i.e., helping peers, showing concern for distress, showing kindness, cooperative, and having concern for moral issues). The measure was designed to assess peer-related social competence among typically developing children and those demonstrating peer rejection and social difficulties. Past work using this measure with children with ASD has demonstrated convergent validity with indices of self-regulation and social-communication (Jahromi et al., 2013). Cronbach’s alpha for the present study was .84.

Results

Analytic Plan

We assessed group differences between children with ASD and those with typical development in each of the emotion prerequisites using ANCOVAs. Mental age was included as a covariate to account for the fact that, although ASD and TD groups were matched on mental age, there was nevertheless variability in children’s mental age that was related to measures of emotion knowledge (rs ranged from .35 to .55, ps ranged from .04 to < .001).

Next, we explored correlations among study variables to understand patterns of association among the emotional competence prerequisites, and between indices of emotional competence and children’s social competence. Give the small sample sizes within groups (n = 20), Kendall’s tau non-parametric correlations were conducted (Bonett & Wright, 2000; Walker, 2016). Kendall’s tau is less sensitive to error (i.e., more robust to outliers) and the obtained p-values for Kendall’s tau are more accurate with smaller samples than other tests of correlation.

Emotion Knowledge

Differences were tested between children with ASD and those with typical development for each of the measures of emotional knowledge. No significant differences were found in children’s expressive or receptive affective labeling, or in children’s stereotypical affective perspective taking responses. Importantly, a significant group difference emerged for non-stereotypical affective perspective taking, F(1,38) = 7.83, p = .01, Cohen’s d = .91, indicating a large effect (Cohen, 1988). Children with ASD (M = 1.12, SD = .72) demonstrated lower affective knowledge on the non-stereotypical perspective-taking component of the task than did those with typical development (M = 1.66, SD = .44). See Table 1 for descriptive and group difference data on study variables.

Coping Strategies With Peers

Next, differences were tested between children with ASD and those with typical development in their use of all emotion regulation coping strategies. A significant difference was found in seeking support, F (1, 38) = 7.32, p=.01, Cohen’s d = 1.02, (M_ASD = 3.48, SD_ASD = 1.69 and M_TYP = 4.72, SD_TYP=1.12), cognitive restructuring, F (1, 38) = 12.21, p =.001, Cohen’s d = 1.09 (M_ASD = 1.38, SD_ASD = .67 and M_TYP = 2.30, SD_TYP = .98), and instrumental coping, F (1, 38) = 17.63, p < .001, Cohen’s d = 1.33 (M_ASD = 2.17, SD_ASD = .89 and M_TYP = 3.56, SD_TYP = 1.17). The above-mentioned strategies were used significantly more frequently by typically developing children (large effect sizes were observed). Finally, a trend difference emerged in children’s reported use of aggressive coping, F (1, 38) = 4.07, p = .05, Cohen’s d = .64 (M_ASD = 3.57, SD_ASD = 1.74 and M_TYP = 2.65, SD_TYP = 1.05); children with ASD used marginally more of this strategy (a medium effect size was observed). No group differences were found in the use of distraction/avoidance strategies or in the use of venting. Based on the above findings, a composite measure was created to reflect greater use of positive forms of coping (e.g., cognitive restructuring, seeking support, and instrumental coping) and less use of negative forms of coping (e.g., aggressive) for use in subsequent correlational analyses. The composite measure showed significant group differences, even after controlling for mental age, F (1, 38) = 21.33, p < .001, Cohen’s d = 1.41 (large effect) such that children with typical development showed more positive coping strategies (M = 4.37, SD = .62) than children with ASD (M = 3.29, SD = .88).

Emotion Regulation, Negative Temperament, and Prosocial Behavior

Children with ASD were rated significantly lower in the composite measure of emotion regulation than their typically developing peers, F (1, 38) = 9.26, p = .004, Cohen’s d = .96 (large effect size), which including significantly lower scores on the regulation subscale, F (1, 38) = 8.33, p = .006, Cohen’s d = .93 (large effect size), and significantly higher scores on the lability/negativity subscale, F (1, 38) = 6.52, p = .015, Cohen’s d = .83 (large effect size). Children with ASD did not differ from those with typical development in negative temperament. Finally, children with ASD were rated as having significantly lower prosocial behaviors with their peers as compared to children with typical development, F (1, 38) = 5.65, p = .02, Cohen’s d = .78 (medium effect size).

Associations Among Emotional Prerequisites and Children’s Prosocial Behavior

Intercorrelations Among Emotion Prerequisites

Kendall’s tau correlations were calculated to explore the associations among the emotional prerequisite variables, and between prosocial behavior and emotional prerequisites. See Table 2 for all correlation analyses. Among children with ASD, results revealed positive associations between their non-stereotypical affective perspective taking and the use of coping strategies with peers. Children with greater non-stereotypical emotion knowledge were rated as more likely to use venting, support seeking, cognitive restructuring, and instrumental coping strategies in the context of emotionally charged peer interactions. Children with positive coping strategies in the context of peers were reported to have higher emotion regulation.

Table 2.

Kendall’s Tau Correlations Among Study Variables Within the ASD (n = 20) and TD (n = 20) Groups

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17
1. Knowledge Happy	—	.55 *	.41 +	.22	.30	.32	−.12	−.08	.25	.05	.24	−.21	−.14	.29	.29	−.11	.05
2. Knowledge Sad	.82 **	—	.86 **	.58 **	.49 *	.53 *	.17	−.18	.13	.05	.29	−.04	.20	.43 *	.26	.04	.25
3. Knowledge Angry	.57 *	.68 **	—	.58 **	.51 **	.34	.13	−.13	.10	.12	.26	−.09	.11	.23	.28	−.03	.11
4. Knowledge Scared	.47 *	.60 **	.64 *	—	.32	.52 **	.30	.26	−.16	.04	.21	.17	−.04	.12	−.06	−.01	.22
5. Knowledge Stereotyp.	.45 *	.53 *	.26	.25	—	.09	.15	−.09	−.18	−.08	.23	−.06	−.10	−.07	.12	−.24	−.10
6. Knowledge Nonstereo.	.40 *	.46 *	.63 **	.55 **	.58 **	—	.23	.07	−.02	−.20	.20	.14	.13	.34 +	−.21	−.03	.16
7. Coping Strategies	.17	.24	.24	.22	.29	.37 *	—	.10	−.34*	.12	.55 **	.33 *	.31 +	.29 +	−.07	.24	.26
8. Avoidance Coping	−.11	−.25	−.19	−.10	.16	.07	.05	—	.31 +	.15	−.05	.32 +	−.22	−.19	−.28	−.07	.19
9. Aggressive Coping	.20	.20	.33	.10	−.11	.03	−.37*	−.20	—	.11	−.03	.32 +	.26	.26	.44 **	.27	.09
10. Venting Coping	.39 *	.38 *	.40 *	.20	.19	.34 *	.26	−.09	.25	—	.21	−.01	.02	.00	.23	.46 **	.25
11. Support Seek Coping	.33 +	.38 *	.50 **	.23	.34	.51 *	.58 **	.02	.03	.51 **	—	−.06	.34 *	.48 *	.22	.24	.19
12. Cognitive Restructure Coping	.22	.27	.42 *	.47 *	.00	.31 +	.20	.11	.32 +	.34 +	.33	—	.01	−.05	−.13	−.04	.39 *
13. Instrumental Coping	.21	.30	.51 *	.45 *	.19	.56 *	.52 **	.09	.09	.42 *	.63 **	.57 **	—	.58 **	.13	.42 *	.31 +
14. Emotion Regulation	.28	.30	.27	.18	.12	.24	.50 **	.00	−.12	.53 **	.53 **	.22	.42 *	—	.18	.18	.32 +
15. Emotion Lability	.21	.14	.18	.12	−.04	.10	−.23	−.23	.59 **	.34 *	.04	.14	.11	−.03	—	.32 +	.02
16. Negative Temp.	.169	.18	.29	.08	.12	.19	−.24	−.17	.66 **	.39 *	.09	.24	.14	−.02	.56 **	—	.21
17. Prosocial w Peers	.30	.39 +	.44 *	.17	.09	.15	.45 *	−.15	−.04	.45 *	.38 *	.25	.36 *	.61 **	−.14	−.03	—

⁺p < .10;

^*p < .05;

^**p < .01.

Values represent Kendall’s Tau correlations. Correlations for ASD group are below the diagonal and those for TD group are above the diagonal.

Among typically developing children, non-stereotypical emotion knowledge and knowledge of sad emotions was positively related to emotion regulation. Those children who were rated higher in emotion regulation were more likely to demonstrate support-seeking and instrumental coping strategies in the context of peers, and those rated high in lability/negativity were more likely to use aggressive coping strategies and to be rated higher in temperamental negativity. Finally, children rated higher in temperamental negativity were more likely to use venting and instrumental coping strategies (See Table 2).

Associations Between Emotional Prerequisites and Social Competence

For children with ASD, prosocial behaviors with peers was positive related to emotion regulation, and specific forms of coping strategies, including venting, support-seeking, and instrumental strategies. In terms of specific emotion labels, knowledge of angry emotions was positively related to prosocial behavior with peers among children with ASD. Finally, among typically developing children, prosocial behaviors with peers was positively related to emotion regulation and individual forms of coping, including cognitive restructuring and instrumental coping strategies (see Table 2).

Discussion

For children with ASD, disrupted peer relationships can be a source of significant challenge. This study makes an important contribution to our understanding of associations between multiple elements of emotional competence and positive prosocial behaviors for these children. Our study revealed several important patterns of findings for this group. Children’s emotion knowledge, especially that which reflects an understanding of others’ non-stereotypical emotional experiences, and their emotion regulation showed a consistent pattern of associations with other aspects of emotional competence. Moreover, use of discrete coping strategies in the context of peers was associated with children’s prosocial peer behavior one year later; for children with ASD, venting, support-seeking, and instrumental coping were all positively related to prosocial behaviors with peers. For typically developing preschoolers, cognitive restructuring and instrumental coping showed positive associations with peer social competence.

Our finding of a significant difference between children with ASD and their typical peers in non-stereotypical emotion knowledge (i.e., when the story character’s emotion was different than that which the child him/herself would have felt in that situation) was consistent with past work indicating challenges in ambiguous, complex, and social emotions among this population (e.g., Bal et al., 2010; Capps et al., 1992; Heery, et al., 2003; Williams & Happé, 2010). Some scholars have argued that because children with ASD have a more limited and scripted repertoire of complex emotions, they may rarely take the perspective of others and have difficulty applying their emotional knowledge to different social situations (Kasari et al., 2001). It may be that the difficulty children with ASD face with non-stereotypical emotional knowledge could be a function of their delays in theory of mind, which is defined as children’s ability to consider multiple perspectives and attribute different mental states (e.g., beliefs and emotions) to others (Ashwin et al., 2006; Baron-Cohen, 2000). Although such abilities develop in neurotypical children by around 4 years of age, children with ASD show continued challenges with these skills (Baron-Cohen et al., 2000). Overall, these findings are consistent with Chamberlain et al.’s (2007) conclusions that children with ASD need a great deal of support in strengthening their repertoire of emotion knowledge and linking such knowledge to different social situations; as will be discussed below, social skills intervention efforts have been developed for children with ASD, aimed at targeting both theory of mind and emotion knowledge (e.g., Bauminger, 2007; Young & Posselt, 2012). Finally, in line with the notion that knowledge of negative emotions may be particularly important to social competence among children with ASD (e.g., Boraston, et al., 2007; Wallace et al., 2011) and the present study’s finding that knowledge of anger among children with ASD was associated with prosocial behaviors in peers, future intervention work may be particularly effective if there is an emphasis on negative emotions.

Our study also revealed important group differences in the coping strategies used by children in the context of peer interactions, which are consistent with findings from laboratory tasks and extend our knowledge of the degree to which children with ASD are able to rely on more sophisticated means to regulate their distress in the context of peers. Specifically, we found that children with ASD were reported to use less instrumental coping, cognitive restructuring, and support seeking than their mental age and language age-matched peers, whereas children with ASD used marginally more aggressive coping strategies. In the lab setting, when faced with a toy removal, children with ASD were shown to use fewer adaptive strategies (e.g., engaging in an alternate activity) but more defending behavior (Konstantareas & Stewart, 2006). Similarly, when faced with frustration (i.e., locked box and impossible puzzles tasks), children with ASD used fewer constructive strategies than their typical peers; support-seeking behaviors were not found to be effective (i.e., not likely to be temporally associated with a subsequent decrease in negativity) for children with ASD, despite being effective for typically developing children (Jahromi et al., 2012). The fact that children with ASD in our study were reported to use less support-seeking strategies when faced with challenging peer situations may reflect a disinclination among these children to turn to others for instrumental support or poorer planning skills (i.e., inability to foresee the potential that an adult may effectively solve the social situation). What is particularly interesting is the fact that, in a study of adolescents with ASD, Rieffe et al. (2011) found that adolescents with ASD also self-reported less use of social support-seeking as a means of coping with a problem as compared to their typical peers. We should highlight that our measure of coping strategies relied on parent report, and therefore findings should be confirmed with future work using observations of children’s behaviors with peers. What also remains to be explored are other developmental skills that are associated with children’s coping strategy use (e.g., planfulness and flexibility) as well individual differences in the effective deployment of coping strategies. It may be, for example, that underlying attentional skills influence children’s strategy-use by aiding in the monitoring of emotion arousal. Future work should continue to explore these children’s use of coping strategies, as there appear to be continued challenges selecting adaptive strategies and/or inhibiting maladaptive strategies even into adolescence (Mazefsky et al., 2014).

An interesting pattern of findings emerged regarding children’s negative emotional expressions. First, we found no differences between groups in temperamental negativity, consistent with past work on preschoolers with ASD using the CBQ measure (Konstantareas & Stewart, 2006). However, children with ASD in our study did have lower scores on lability/negativity from the measure of emotion regulation. It may be that, in general, preschoolers overall negative affect in the context of day-to-day activities, like those measured in the temperament measure, are not significantly different than those of their typical peers, but that intense, uncontrolled emotions and outbursts are more likely to occur for children with ASD, reflecting these children’s dysregulation in emotionally arousing situations. Importantly, lability/negativity was also related to children’s greater use of aggressive coping strategies for both groups of children and venting strategies for children with ASD. The consequence of heightened lability/negativity may be that it undermines healthy coping strategies that ultimately promote positive interactions with peers.

Interestingly, although negative affectivity was not significantly related to prosocial behaviors, there was a positive association between emotion regulation and prosocial behavior in both the ASD and TD groups. This finding is similar to that of Denham et al. (2003), who found emotion regulation to be a more salient predictor of social competence than emotional expressiveness, and who argued that children’s enduring moods and patterns of negativity may be less important than how they respond to these experience (i.e., with regulated versus dysregulated responses). Because it has been argued that negative affect may not only inhibit children’s own prosocial behaviors but also evoke fewer positive initiations from peers (Denham et al., 2003), it may be that such behaviors are more closely related to rejection of children of ASD by their typical peers rather than their own initiations of prosocial interactions. Although our study offers a first step in understanding how these emotional factors may be linked to positive social outcomes of children with ASD, future work should examine these behaviors using observations.

Implications

Most intervention efforts targeting children’s self-regulation incorporate the principles of cognitive-behavior therapy (CBT), which emphasizes children’s awareness or recognition of their emotional reactions while strengthening their capacities to problem solve and achieve positive social and emotional outcomes. CBT has been found effective for anxiety in adolescents with ASD (Wood et al., 2009) and when targeting emotional reactions (anger and anxiety) and coping strategies (Scarpa & Reyes, 2011). Other approaches have addressed the social challenges of children with ASD by targeting underlying emotion recognition together with social-cognitive skills (e.g., theory of mind), with reported improvements in children’s ability to identify components of facial expressions (Ryan & Charragain, 2010) as well as in theory of mind performance (Stichter et al., 2010). An example of such an approach is Ratcliffe and colleague’s (2014) Emotional Based Social Skills Training (EBSST), which incorporates understanding one’s own and others’ emotions, emotional problem solving or coping strategies, and emotion regulation skills in one-on-one counseling settings within school as well as trainings at home with parents. Positive effects on social behaviors were also demonstrated by Young and Posselt (2012), who incorporated a strong emotion understanding component in their social skills training DVDs called “The Transporters”, in which cartoon character vehicles teach lessons on understanding faces and matching facial expressions with felt emotions. It is noteworthy that, when Williams et al., 2012 extended this approach to children with ASD with lower levels of cognitive functioning, children’s improvements were limited to the recognition of anger emotions, and the effects were not maintained at follow-up. This indicates the need for future social-emotional interventions for youth with ASD to focus on children with lower levels of cognitive functioning and/or minimally verbal children to fully assess whether current approaches can be generalized to the broader population. Importantly, Bauminger (2002; 2007) has found that children with ASD can show improvements in social problem solving in the context of peers alongside emotion knowledge, including understanding of complex emotions (Bauminger 2002; 2007; Sticher et al., 2010; Solomon et al., 2004), and results have been reported to have been maintained at follow-up (Bauminger, 2007). Given the link between emotion recognition and children’s capacity to regulate their emotion, the above-mentioned work is a promising approach to improve children’s self-regulatory challenges. The direct instruction of emotional knowledge is a key component of these interventions that has shown promising results, but it may be that the application of the skills associated with social competence is what may need further development for many children. Bauminger’s (2007) work is notable, in that she incorporates practice with peers which may be important for encouraging the application of adaptive strategies. Future work should continue to take such curriculum out of the one-on-one (e.g., counseling) setting and into the classroom or other group setting, with peers so that everyone gains access to the same knowledge and opportunities to practice the strategies using the real-life, nonstereotypical and often-nuanced types of social situations that tend to occur classroomwide.

Limitations and Future Directions

The study had several limitations that may point to directions for future research. First, our sample of preschoolers with ASD did not have significant language or cognitive delays, thus our findings may not generalize to the broader population of children with ASD, including those with intellectual disabilities and/or significant language delays. Moreover, while we utilized a research-reliable ADI-R to confirm children’s diagnoses, the study would have been strengthened by including an additional observational diagnostic measure like the ADOS-2. Additionally, while we used a combination of direct child measures (i.e., to measure emotion knowledge) and parent report, it will be important for future research to continue to examine these constructs with a wider range of measures and carefully controlled samples; such work with help to further elucidate many of the mixed findings resulting from this line of research. Moreover, given the often-nuanced aspects of social interactions that result in challenges for children with ASD, it will be important for future work to measure emotional and social capacities in real-life settings with peers using observational measures. Our sample was carefully controlled and matched on key elements of children’s developmental functioning, including mental age, expressive language, and receptive language. That said, the relatively small size of our sample may have undermined some of our findings. As such, our work is a first step toward understanding these processes in children with and without ASD, and future work should aim to test these types of constructs in larger samples, and using longitudinal models (Denham et al., 2003). Finally, future work should also explore the role of gender in these relations. For example, Denham et al. (2013) found that emotion regulation was a stronger predictor of social competence for typically developing girls than boys. Because of the predominance of boys within the ASD population, most studies like ours have too few, if any, girls to test such assumptions. An effort should be made in future research to better understand whether patterns of emotional competence and social competence are similar for girls with ASD.

In conclusion, this study makes an important contribution to our understanding of how different components of emotional competence are linked to social competencies among children with ASD and their typical peers. Our findings also point to the importance of focusing on key dimensions of emotion knowledge (non-stereotypical affective perspective taking) for future research. It will be important for future intervention research to continue to explore the role of underlying skills that promote these children’s social competence, while also identifying ways to intervene in situ, such that interventionist can model and practice social-emotional skills in the presence of children with ASD and their typical peers, together, to maximize the potential generalization of such knowledge and skill development.