Cognitive Attributions and Emotional Expectancies Predict Emotions in Mother-Adolescent Interactions

Abstract

The purpose of this study was to examine adolescent’s hostile attributions of mother’s intent and emotional self-expectancies as contributors to expression of emotion between mothers and adolescents. Data were collected from 268 10- to 12-year-olds (133 girls, 135 boys) and their mothers. Each dyad was observed in a conversational activity that was coded for both partners’ expressions of four discrete emotions: happiness, anger, sadness, and fear. Adolescents responded to hypothetical stories to assess their emotional expectancies and attributions. Regression analysis revealed that adolescents’ hostile attributions and emotional expectancies made independent contributions to the expression of happiness and anger with their mother. Adolescents who interpreted their mother’s hypothetical behavior as hostile, and who expected to feel less happiness and more anger in response to their mother’s hypothetical behavior, expressed less happiness and more anger with their mother. The findings support the conceptual distinction between hostile attributions and emotional expectancies.

Keywords

mother-adolescent emotional expectancies hostile attributions expression of emotion

Socialization theories based on information processing models identify attributions as playing a key role in the quality of parent-child relationships (Bugental & Johnston, 2000; Creswell, Murray, & Cooper, 2010). Attributions are interpretations individuals make about the behavior of others and the reasons that particular social events take place so as to construct goals and formulate responses in the context of interactive exchanges (Weiner & Graham, 1984). Attributions can take many forms, but attributions of intent made about a partner’s behavior appear to be robust predictors of relationship quality (Brody, Arias, & Fincham, 1996; Yeager, Miu, Powers, & Dweck, 2013). Hostile attributions of intent, in particular, have been identified as a deficit in interpreting social cues that contributes to interpersonal difficulties (Reijntjes et al., 2011). In his multistep social information processing model of interpersonal interactions, Dodge (2006) suggests that for some children an attributional style of inferring hostile intent in response to nonhostile and ambiguous social events (i.e., hostile attribution bias) becomes a personality-like characteristic that endures across time and guides behavior. Consistent with this perspective, empirical evidence suggests that individual differences across children in hostile attribution style predict poor quality parent-child relationships (MacKinnon-Lewis, Lamb, Hattie, & Baradaran, 2001; MacKinnon-Lewis et al., 1994).

Although considerable empirical support has been garnered for social information processes outlined by Crick and Dodge (1994), a major criticism leveled against the model is that it fails to consider emotional processes that coincide with cognitive appraisals (Izard, 2001; Izard, Fine, Mostow, Trentacosta, & Campbell, 2002). Theorists have argued that emotions perform a critical role in how children organize, prioritize, and interpret salient social information (Izard et al., 2002; Lemerise & Arsenio, 2000). The part that emotions play during the initial step of interpretation may be particularly important in leading to subsequent errors in social information processing, such as hostile attribution biases. In this initial step, children access latent mental structures, or social knowledge, in response to ambiguous or complex social cues (Crick & Dodge, 1994). Children use these knowledge structures that act as implicit, automatic memories of past experiences (Bugental & Johnston, 2000; Burks, Laird, Dodge, Pettit, & Bates, 1999) to assist in filling in missing social information (Burks et al., 1999; Crick & Dodge, 1994). Such knowledge structures contain both affective and social-cognitive information that children subsequently use to interpret situational cues (Lemerise & Arsenio, 2000). Consequently, understanding the nature of links between affective and social-cognitive processes may yield insight into the origins of children’s socioemotional behavior.

One form of emotional knowledge structure that has been identified by empirical studies is emotional expectancies (Arsenio, Gold, & Adams, 2004; Krettenauer, Jia, & Mosleh, 2011). Also referred to as emotion attributions (Arsenio & Lemerise, 2004) and emotional schemata (Izard et al., 2002), emotional expectancies have been defined as beliefs regarding the emotions that are likely to be experienced during some future event (Harris, 1985). Theoretical accounts of emotional expectancies argue that emotional antecedents and consequences of social events are highly salient and likely to be stored in long-term memory (Lemerise & Arsenio, 2000; Teasdale, 2005). When subsequently triggered by environmental cues, these memories are used to form emotional expectancies to anticipate likely responses to new situations and to plan potential behaviors (Arsenio, Gold, & Adams, 2004; Krettenauer et al., 2011). In this way, self-evaluative emotional expectancies are considered to serve as a link between past experience and current behaviors (Lemerise & Arsenio, 2000; Teasdale, 2005). Consequently, emotional expectancies may be especially relevant to the emotions that individuals express in social situations.

The links between children’s emotional expectancies and emotions expressed in social situations may be particularly evident during interactions with their mother. Moreover, connections between emotional-cognitive processes and emotions are likely to be highly salient during the transition to adolescence. Evidence suggests that as children enter into puberty the number and variety of emotion-cognition connections that relate to their self-concept and social life rapidly increase (Krettenauer et al., 2011; Teasdale, 2005). At the same time, negative emotion tends to increase during adolescence in relation to experiences within the family (Chung, Flook, & Fuligni, 2009; Shanahan, McHale, Osgood, & Crouter, 2007). Early adolescence is also a time when peers become an increasingly important source of excitement and enjoyment, whereas the parent’s role in fulfilling a child’s emotional needs diminishes (Larson, Richards, Moneta, Holmbeck, & Duckett, 1996). Perhaps because of this, many children appear to experience a qualitative change in their relationship with their mother during early adolescence that is characterized by a slight increase in negative emotions and a decrease in maternal warmth and uncritical acceptance of the child (Collins & Laursen, 2006; Shearer, Crouter, & McHale, 2005). Individual differences in the quality of emotions displayed between mother and child during the transition to adolescence may be accounted for by the particular content of youth’s attributions of mother’s behavior and emotional expectancies, as well as the interaction between these two forms of social-cognitive structures.

Theorists explain the formation of connections between cognitive processes and emotion as a feedback loop (Krettenauer et al., 2011; Lewis, 2000; Teasdale, 2005). For example, Lewis (2000) argues that positive feedback between emotions and cognitive appraisals is the basis for self-organizing interpretations of events, and the foundation for personality patterns. According to this view, emotion guides an individual’s attention to particular goal-relevant elements in a situation. The individual forms an appraisal that generates further emotion, that is, in turn, fed back into the system through repeated iterations. The resulting cognition-emotion structures are considered to be the fundamental psychological elements that work together to form macropersonality structures (Lewis, 2000; Magai & McFadden, 1995) and guide individual styles of processing information and engaging with the world (Izard, 1977).

These theoretical explanations of the feedback processes involved in children’s formation of cognitive-emotional structures suggest that individual differences in the content of emotional expectancies may be related to variations in adolescents’ expressions of emotion in social interactions. For instance, Crick and Dodge (1994) point out that children’s internal emotional states together with other situational cues serve as important sources of information to be encoded and interpreted in social situations. In this way, an adolescent’s mood and emotions may affect what is noticed about a social encounter and may be included as part of the cognitive information that is stored for later reference. In subsequent social encounters, children’s cognitive-emotional structures may be triggered in such a way as to make the recollection of mood-congruent information more likely, thus influencing the adolescent’s interpretation of ongoing social interaction and giving rise to the expression of particular emotions.

Considered from this framework of attribution theory, individual differences in the quality of mother-adolescent emotions may be accounted for by adolescents’ attributions, emotional expectancies, and the interaction between these two knowledge structures. Specifically, adolescents who hold hostile attributions are more likely to express negative emotion and less positive emotion when interacting with their mother. Likewise, teens who enter into an interaction with their mother expecting to become angry are more likely to express anger. Children who hold both hostile attributions and negative emotional expectancies may be particularly likely to express negative emotion, and less positive emotion, than children who have hostile attributions but positive emotional expectancies, or nonhostile attributions and negative expectations. Because no study has examined both forms of knowledge structure simultaneously, it remains unclear to what extent hostile attributions and emotion expectancies make unique or overlapping contributions to adolescents’ expressions of emotion with their mother.

The purpose of this study was to explore associations between early adolescents’ attributions, emotional expectancies, and the emotions expressed by youth when interacting with their mother. A major focus was to examine individual differences in the manifestation of hostile attributions and emotional expectancies. Based on the review of existing literature, four hypotheses were examined in relation to the association between adolescents’ hostile attributions and expression of emotion with mothers. Specifically, it was hypothesized that adolescents’ hostile attributions would be associated with low levels of happiness (Hypothesis 1a [H1a]), high levels of anger (Hypothesis 1b [H1b]), high levels of sadness (Hypothesis 1c [H1c]), and high levels of fear (Hypothesis 1d [H1d]). Similarly, four hypotheses were examined in relation to the association between adolescents’ emotional expectancies and expression of emotion with mothers. Specifically, it was hypothesized that emotional expectancies of happiness would be associated with the expression of happiness (Hypothesis 2a [H2a]), emotional expectancies of anger would be associated with the expression of anger (Hypothesis 2b [H2b]), emotional expectancies of sadness would be associated with the expression of sadness (Hypothesis 2c [H2c]), and emotional expectancies of fear would be associated with the expression of fear (Hypothesis 2d [H2d]). It was also of interest to examine how adolescents’ hostile attributions and emotional expectancies may interact in relation to adolescents’ expressions of emotion. However, no specific hypotheses were advanced regarding potential interaction effects given the lack of empirical evidence with which to guide such formulations. Likewise, it was of interest to explore the associations between adolescent’s emotional expectancies and mother’s expression of emotion. Again, however, due to the paucity of past research on which to propose possible correlations no specific hypotheses were formulated.

Method

Participants

Study participants were 268 mother-adolescent dyads who were part of a larger longitudinal study investigating adolescent’s adjustment to middle school. Participants were recruited from 13 primary schools in a midsized town (approximately 197,000 inhabitants) in the southeast United States. Using rosters and demographic information concerning all children enrolled in the fifth grade obtained from school administrators, researchers made initial telephone contact with families during the summer before the target child entered middle school.

There were 300 families (72% of those contacted who were eligible) who agreed to participate and attended the first wave of data collection (T1) during the summer prior to entering middle school. Approximately 4 months later, during the middle of the school year, 281 families returned to the laboratory for the second wave of data collection (T2). The third wave of data collection (T3) was carried out during the summer following the first year of middle school, approximately 6 months after T2 and 12 months after T1. At T3, 268 families, comprising 114 African American and 154 European American families, with 133 girls and 133 boys, returned to the laboratory. This was the final sample for the present study. As a condition for participation all mothers were married, with length of marriage ranging from 15 to 27 years $\bar{X}$ = 18 years). Total family income ranged from US$16,400 to US$127,560 $\bar{X}$ = US$55,630). Of the mothers, 14% had less than a high school education, 31% had received a high school diploma, and 28% had completed an associate of arts or bachelor’s degree. Analysis comparing the group of families who completed all three phases of data collection (n = 268) with those who dropped from the study (n = 32) revealed no significant differences on any demographic characteristics, χ²(1, 300) range = 0.81 to 2.11, ns, or continuous variables, t(300) range = 0.12 to 1.45, ns, used in the present study.

Procedure

During the summer prior to children’s entry into sixth grade, mother-child dyads were invited to a research laboratory located on the campus of a midsized University. Following procedures approved by the university Institutional Review Board, at the visit mothers provided consent for their own and their child’s participation in the study, and children provided assent for their own participation. Mothers and children were taken to separate rooms to complete questionnaires on their own, with a researcher reading the survey to the child to ensure that questions were answered correctly and that the laboratory visit proceeded in a timely manner. Approximately 4 months later, during the middle of children’s sixth grade year, mother-child dyads were contacted again and invited to return to the laboratory for a follow-up visit. Finally, during the summer following children’s matriculation from sixth grade, mother-child dyads were invited for a third visit to the laboratory for a follow-up visit. The exact same data collection procedures were used by researchers during T1 and T3. During T2 mothers and children completed questions concerning children’s adjustment to school.

At both T1 and T3, mothers and children completed questionnaires pertaining to family demographics, child-rearing practices, child adjustment, and family dynamics. In addition, each mother-child dyad participated in a 20-minute discussion task developed by Conger and his colleagues (Conger et al., 1993; Conger et al., 1991). For the purpose of this study, only mother’s report of family demographic data and observational codes from the 20-minute discussion task are used, thus other measures collected at T1 and T3 are not described further.

As part of the discussion task at T1 and T3, mother and child responded to 14 questions about family life presented on index cards, beginning with relatively innocuous topics (e.g., first question, “What do I do with my mother when we spend time together?”) and gradually moving to more provocative issues (e.g., 14th question, “What do my parents usually do when I get into trouble for something?”). Mother-child dyads were told that the cards had identifying labels indicating whether they were to be read aloud by the mother or by the child (e.g., Card 4, read by the child, “What are some rules or things my Mom expects me to do or not do? Which of these are fair and which are unfair?” Card 12, read by the mother, “If each of us could change anything about our family, what would we like to change? Why? Do we agree or disagree about this?”). Mother and child were instructed to discuss their responses with each other and to proceed through the cards in consecutive order. To ensure that the mother and child understood the task, the researcher had the dyad complete a practice card and provided follow-up instructions as necessary. Mother-child dyads were videotaped from behind a one-way mirror for later coding. The majority of mother-child dyads (88%) completed discussion of all 14 cards within the 20 minutes allotted.

At T2, mothers completed questionnaires pertaining to children’s adjustment to school. Children completed questionnaires pertaining to their perceptions of school climate, the quality of their peer relationships, and were interviewed using the Child Attribution Measure (MacKinnon-Lewis, Lamb, Arbuckles, Baradaran, & Volling, 1992). For the purpose of this study, only children’s responses to the Child Attribution Measure were included in the data, therefore other measures collected at T2 are not described further.

Measures

Family demographics

Mothers provided information concerning family demographics by completing the Family History Inventory (MacKinnon-Lewis et al., 1992) at T1. The Family History Inventory contains 14 questions that include annual household income, parent age, parental education, and the number of children in their household.

Mother-child expression of emotion

A microanalytic coding system was used to code the videotapes of the mother-child interaction session at both T1 and T3. The coding system was designed based on theoretical conceptualizations of emotion expression occurring across multiple modalities of facial expression, verbal intonation, and nonverbal behavior (Izard, 1977; Magai & McFadden, 1995). For each 15-second interval of interaction, coders noted for both mother and child the presence (1) or absence (0) of four discrete emotions. Happiness was coded as present if the target displayed positive physical gestures, such as smiling, laughing, kissing, hugging, and/or statements indicating approval, or affection toward the partner (e.g., “good job!”). Anger was coded as present if the target displayed instances of whining, nagging, complaints, or sarcasm, accompanied by negative affect, as well as disruptive behavior such as yelling or name-calling and physical aggression (e.g., pushing, excessive roughness). Sadness was coded as present if the target displayed crying, sorrow, grief, melancholy, or dejection in their facial expressions, body language, or tone of voice, as well as statements reflecting sadness or sorrow (e.g., “I want to cry when I hear that.”). Anxiety/fear was coded as present if the target displayed nervousness, worry, apprehension, panic, or alarm in their facial expressions, body language or tone of voice, as well as statements reflecting anxiety or fear (e.g., “I am really worried about you.”). The coding system was developed and reliably used in past research, with evidence of validity provided by significant correlations between the observed emotion scores and parent and teacher ratings of children’s positive, angry, fearful, and sad temperamental tendencies (MacKinnon-Lewis et al., 1992; MacKinnon-Lewis et al., 1994).

A separate score for each of the four emotions, (a) happiness, (b) anger, (c) sadness, and (d) anxiety/fear, was created for both mother and child, based on the number of intervals in which the particular type of emotion was present out of the total number of intervals in the discussion task session. These proportion scores were transformed (i.e., arcsine) to reduce the skew of the data and to better approximate a normal distribution (Cohen & Cohen, 1975). Analyses comparing dyads that completed the discussion task (n = 236) with those that did not complete the task in the 20-minute allotted (n = 32) revealed no significant differences on any of the emotion scores t(268) range = .43 to 1.14, ns. Therefore, no adjustments were made to the data based on task completion status prior to use in analyses.

All coders participated in extensive training, including reviewing the coding manual, observing tapes with a trainer, and coding training tapes until reaching a reliability criteria of κ = .85 with the first author. Once reliable, coders were assigned a set of tapes to code the emotional expressions of 1 partner in the dyad. In this way, no coder was assigned to code both a mother and child from the same dyad. During the course of coding the tapes assigned to them, coders “realigned” after coding every 10 tapes to prevent observer drift. This realignment was accomplished by having all coders participate in a group meeting to review the coding manual and coding the same mother-child dyad for the purpose of comparing scores across all coders. Reliability for the final coding of adolescent expression of emotion was κ = .87 and κ = .89, for T1 and T3, respectively. For mother expression of emotion, reliability was κ = .91 and κ = .88, for T1 and T3, respectively.

Hostile attributions and emotional expectancies

At T2, adolescents completed the Child Attribution Measure (MacKinnon-Lewis et al., 1992) consisting of six hypothetical vignettes read to the child in an interview format. Each vignette describes a situation where the child is subjected to a negative outcome as a result of an ambiguous behavior by the mother. An example of a story is as follows: “Pretend that you and your mom are shopping at the grocery store and that you reach for a candy bar that you want to look at. Your mother tells you that you cannot have it.” After each story, the child was asked three questions: (a) “Why do you think your mother would do this?” (open-ended), (b) “How would that make you feel?” (open-ended), and (c) “Was your mother being nice or mean?” (close-ended). The phrasing and response options were chosen based on the results of previous pilot testing of preadolescents who were demographically similar to the study sample (MacKinnon-Lewis et al., 1992; MacKinnon-Lewis et al., 1994).

Adolescents’ responses to the first open-ended question, “Why do you think your mother would do this?” intended to tap attributions of intent were coded on the following scale: 5 = hostile, the mother acted purposefully with negative intent (e.g., “She was being mean.”; “She was punishing me.”), 4 = moderately hostile (e.g., “She’s didn’t think I deserved it.”; “Because she wanted to use all the money for herself.”), 3 = neutral, the mother acted with harmless intentions or the outcome was accidental (e.g., “She was in a hurry.”; “She didn’t have enough money.”), 2 = moderately prosocial (e.g., “She knew we already had it at home.”; “Because I already had a snack.”) and 1 = prosocial, the mother intended to behave in a positive way (e.g., “She thought it wasn’t healthy for me.” “She didn’t want me to spoil my appetite.”). The scores for the responses to the six stories were summed to form a single index of the child’s hostile attribution of mother’s intent–open-ended. The internal consistency of the composite scale was α = .73.

Adolescents’ responses to the second open-ended question, “How would that make you feel?” intended to capture emotional self-expectancies were coded into one of eight categories: (a) happy (e.g., glad, great, good, nice), (b) embarrassed (e.g., self-conscious, humiliated, ashamed, shy), (c) sad (e.g., disappointed, hurt, crushed, down), (d) afraid (e.g., worried, concerned, surprised, anxious), (e) indifferent (e.g., not mad or happy, nothing, wouldn’t care, alright), (f) general negative (e.g., bad, discouraged, upset, not too happy), (g) angry (e.g., mad, mean, aggravated, frustrated) and (h) responsive (e.g., I understand, she had good reason, she knew it would benefit me). Two coders scored 20% of the vignettes and reliability between the coders’ scores was κ = .86. For the purpose of this study we were only interested in the four basic emotion categories: happy, angry, sad, and afraid. Therefore, for each emotion a proportion score was created based on the sum of the responses that were coded in the emotion category divided by the total number of responses. Prior to use in analyses, these proportion scores were Arcsine transformed to reduce the skew of the data (Cohen & Cohen, 1975).

Adolescents’ responses to the third close-ended question, “Was your mother being nice or mean?” designed to get at attributions of mother’s hostile intent were scored using a 5-point Likert-type scale: 5 = really mean, 4 = mean, 3 = not mean or nice, 2 = nice, and 1 = really nice. The adolescent’s ratings across the six stories were averaged to yield a single index of the child’s hostile attribution of mother’s intent–closed-ended. The internal consistency of the composite scale was α = .78.

Correlations revealed that there was a significant association between the scores for Questions 1 and 2, both of which were designed to assess hostile attributions (r = .55, p < .001). Consequently, to provide a more robust and parsimonious measure of adolescent’s hostile attributions, the two scores were averaged to form a single measure of Hostile Attribution of Mother’s Intent. Because the two scores were based on the same scale range, they did not require standardization prior to averaging them together. The internal consistency of the resulting score was α = .75.

Results

Preliminary Analyses

Means and standard deviations of study variables are shown in Table 1. As can be seen, happiness was the most frequent emotion expressed by both adolescents and their mothers at both T1 and T2. Likewise, adolescents reported emotional expectancies of happiness most frequently, followed by anger, sadness, and fear, respectively. Emotional expectancies of sadness and fear were relatively uncommon.

Table 1.

Descriptive Statistics for Mother and Adolescent Emotional Expression, Adolescent Attributions, and Adolescent Emotional Expectancies.

	T1	T2	T3
	$\bar{X}$ (SD)	$\bar{X}$ (SD)	$\bar{X}$ (SD)
Adolescent emotional expression
Happy	.41 (.27)		.32 (.33)
Angry	.16 (.25)		.20 (.23)
Sad	.11 (.17)		.10 (.20)
Fear	.08 (.14)		.06 (.15)
Mother emotional expression
Happy	.44 (.23)		.37 (.26)
Angry	.11 (.20)		.14 (.27)
Sad	.09 (.15)		.07 (.18)
Fear	.07 (.10)		.09 (.14)
Adolescent hostile attributions		14.62 (2.41)
Adolescent emotional expectancies
Happy		.26 (.24)
Angry		.18 (.20)
Sad		.05 (.14)
Fear		.03 (.09)

To examine the possibility of differences based on sex and ethnicity, we performed a MANOVA using a 2 (Male vs. Female) × 2 (European American vs. African American) between-participants design. The result revealed a significant main effect for child sex on both adolescent expression of emotion, Wilks’ λ = .86, F(3, 265) = 5.80, p < .01, η² = .14, and mother expression of emotion, Wilks’ λ = .90, F(3, 265) = 3.32, p < .05, η² = .10, and a significant main effect for ethnicity on adolescent expression of emotion, Wilks’ λ = .93, F(3, 265) = 3.06, p < .05, η² = .09. There was no significant interaction effect between sex and ethnicity Wilks’ λ = 2.13, F(3, 265) = .94, ns. Follow-up analyses revealed that girls expressed more happiness $\bar{X}$ = .48, SD = .23) with their mothers than boys $\bar{X}$ = .34, SD = .32), F(1, 260) = 22.70, p < .001, η² = .15. In addition, mothers of boys expressed more anger $\bar{X}$ = .15, SD = .17) than mothers of girls $\bar{X}$ = .08, SD = .13), F(1, 260) = 13.51, p < .01, η² = .15, whereas mothers of girls expressed more sadness $\bar{X}$ = .14, SD = .18) than mothers of boys, $\bar{X}$ = .07, SD = .09), F(1, 260) = 10.26, p < .05, η² = .09. European American adolescents expressed more happiness $\bar{X}$ = .45, SD = .24) with their mothers than African American adolescents $\bar{X}$ = .38, SD = .30), F(1, 260) = 8.46, p < .05, η² = .08.

Analyses with adolescents’ social-cognitive functioning as the dependent variables revealed a significant main effect for adolescent sex, Wilks’ λ = .81, F(4, 264) = 6.25, p < .01, η² = .16, and ethnicity, Wilks’ λ = .97, F(3, 265) = 3.81, p < .05, η² = .12. Boys had higher hostile attributions scores $\bar{X}$ = 19.45, SD = 2.25) than girls $\bar{X}$ = 10.23, SD = 2.66), F(1, 260) = 27.70, p < .001, η² = .22, and African American adolescents had higher hostile attributions scores $\bar{X}$ = 16.85, SD = 2.57) than European American adolescents $\bar{X}$ = 11.91, SD = 2.36), F(1, 260) = 9.10, p < .05, η² = .08. Girls had higher happiness $\bar{X}$ = .31, SD = .20), F(1, 260) = 18.83, p < .001, η² = .20, and sadness, $\bar{X}$ = .08, SD = .10), F(1, 260) = 20.87, p < .001, η² = .14, emotional expectancies than boys $\bar{X}$ = .20, SD = .28; $\bar{X}$ = .03, SD = .19, boy’s happiness and sadness emotional expectancies, respectively), whereas boys had higher emotional expectancies of anger $\bar{X}$ = .25, SD = .23) than girls $\bar{X}$ = .11, SD = .17), F(1, 260) = 26.43, p < .001, η² = .20.

Correlations among the variables are presented in Table 2. Of primary interest were the correlations between expression of emotion variables and adolescents’ social-cognitive variables. Adolescents who expressed more happiness reported lower hostile attributions, higher emotional expectancies of happiness, and lower emotional expectancies of anger. Adolescents who expressed more anger reported higher hostile attributions, lower emotional expectancies of happiness, higher emotional expectancies of anger, and higher emotional expectancies of sadness. Adolescents who expressed more sadness reported lower emotional expectancies of anger and higher emotional expectancies of sadness. Adolescents who expressed more fear reported lower emotional expectancies of anger.

Table 2.

Correlations Among Adolescents’ Emotional Expression, Mothers’ Emotional Expression, Adolescents’ Attributions, and Adolescents’ Emotional Expectancies at T1 and T3.

	1	2	3	4	5	6	7	8	9	10	11	12	13
Adolescent emotion
1. Happy	—	.16	.13	−.11	.27**	.07	−.21*	−.07	−.21*	.30**	−.22*	.11	−.02
2. Angry	.23*	—	.29**	.12	.13	.23*	−.10	.04	.33**	−.26*	.18*	.13	.05
3. Sad	.11	.31**	—	.20*	−.13	.24*	.18*	.12	−.13	−.11	−.17*	.15	−.08
4. Fear	−.07	.10	.17*	—	−.13	.07	.27**	.10	.12	.16	−.10	.08	.13
Mother emotion
5. Happy	.36**	.17*	−.09	−.13	—	.17*	.12	.14	−.30**	.20*	−.25*	.23*	.06
6. Angry	−.22*	.28**	.20*	.03	.18*	—	.22*	.14	.20*	.12	.17*	.10	−.10
7. Sad	−.18*	−.12	.20*	.30**	.10	.20*	—	−.04	.21*	−.19*	.12	.23*	−.11
8. Fear	−.02	.07	.16	.10	.17*	.16	−.07	—	.10	−.12	−.05	.11	.12
Cognitive-emotional
9. HAI	−.24*	.39**	−.12	.09	−.38**	.22*	−.17*	.07	—	−.23*	.18*	−.11	.09
10. EE–Happy	.34**	−.20*	−.15	.13	.24*	.18*	−.17*	−.13	−.27**	—	−.17*	.11	−.13
11. EE–Angry	−.19*	.22*	−.20*	−.17*	−.27**	.16	.10	−.08	.22*	−.19*	—	.23*	−.10
12. EE–Sad	.14	.17*	.18*	.05	.20*	.20*	.27**	.17*	−.10	.10	.20*	—	.05
13. EE–Fear	−.11	.09	−.02	.13	.07	−.12	−.04	.10	.06	−.17*	−.11	.02	—

Note. Correlations with T1 emotion expressiveness variables are presented above the diagonal and correlations with T3 emotion expressiveness variables are presented below the diagonal. HAI = Hostile Attributions of Intent; EE = Emotional Expectancies.

p < .05. **p < .01. ***p < .001.

Mothers’ expression of happiness was significantly negatively associated with adolescents’ hostile attributions and emotional expectancies of anger, but positively associated with adolescents’ emotional expectancies of happiness and sadness. Mothers’ expression of happiness was positively associated with adolescents’ hostile attributions, emotional expectancies of happiness, anger, and sadness. Mothers’ expression of sadness was negatively associated with adolescents’ hostile attributions and emotional expectancies of happiness, but positively associated with adolescents’ emotional expectancies of sadness. Mothers’ expression of fear was positively associated with adolescents’ emotional expectancies of sadness.

Primary Analysis

Hierarchical multiple regression analyses were used to test the main hypothesis concerning the associations between adolescents’ hostile attributions and emotional expectancies measured at T2 and expression of emotion with their mother measured at T3. The first three steps in each regression were designed to control for demographic characteristics, adolescent expression of emotion at T1, and partner expression of emotion at T3, respectively, in predicting adolescent emotion at T3. Specifically, sex and ethnicity were entered together in the first step. Step 2 contained the variable that represented adolescents’ expression of emotion at T1 that matched the T3 emotion variable being predicted (e.g., in the regression predicting T3 happiness Step 2 contained the adolescents’ happiness at T1; in the regression predicting T3 anger Step 2 contained the adolescents’ anger at T1). Step 3 contained the variable that represented mothers’ expression of emotion at T3 that matched the adolescent emotion variable being predicted (i.e., in the regression predicting adolescent T3 happiness Step 3 contained mothers’ happiness at T3; in the regression predicting adolescent T3 anger Step 3 contained mothers’ anger at T3). Given space considerations, results from Steps 1, 2, and 3 are not described in the text but are included in the tables. Adolescent hostile attribution was entered in Step 4 of each analysis. Adolescent emotional expectancies scores were entered together in Step 5. The product of the respective hostile attributions and emotional expectancies variables was as an interaction in Step 6 (Aiken & West, 1991). In total, four separate regression analyses were conducted for each of the four adolescent expression of emotion variables with mother (happy, angry, sad, fear) at T3.

Calculation of simple intercepts and simple slopes to probe significant interactions was conducted according to standard procedures (Aiken & West, 1991), including centering the variable prior to use in analyses. These analyses yielded intercepts and slopes representing the relations between the predictor (adolescent hostile attributions) and outcome (child expression of emotion) at lower (1 SD) and higher (+1 SD) levels of the moderator (emotional expectancies). It is important to note that a significant interaction term indicates that the association between the predictor and outcome variable differs at higher level of emotional expectancies than at a lower level of emotional expectancies. The significance of the slope itself indicates whether the magnitude of the slope is significantly different from zero at a particular level of emotional expectancies. For those interactions that were significant, as a secondary analysis to more fully test the level of hostile attributions of intent that makes the most difference, we employed the Johnson-Neyman statistical approach to examine the regions of significant confidence intervals (Preacher, Curran, & Bauer, 2006).

Predicting adolescent expression of emotion

Two of the four hypotheses involving the relationship between adolescent hostile attributions and expression of emotion were supported. Specifically, as shown in Table 3, adolescent hostile attributions predicted less happiness (H1a) and more anger (H1b) at T3 even after controlling for adolescent expression of the corresponding emotion at T1 and mother expression of the corresponding emotion at T3. Contrary to hypotheses, adolescent hostile attributions were unrelated to expression of sadness (H1c) or fear (H1d). In contrast, all four of the hypotheses concerning the relationship between adolescent emotional expectancies and expression of emotion were supported. That is, emotional expectancies of happiness predicted the expression of happiness (H2a), emotional expectancies of anger predicted the expression of anger (H2b), emotional expectancies of sadness predicted the expression of sadness (H2c), and emotional expectancies of fear predicted the expression of fear (H2d). Although not hypothesized, there also were some interesting patterns of association between different emotions, with emotional expectancies of sadness being negatively associated with emotional expressions of happiness, emotional expectances of fear being negatively associated with expression of anger, emotional expectancies of happiness being negatively associated with expression of sadness, and emotional expectancies of anger being negatively associated with expression of fear.

Table 3.

Adolescent Hostile Attributions and Emotional Expectancies as Predictors of Adolescents’ Expression of Emotion With Mother at T3.

	Happy (T3)			Angry (T3)			Sad (T3)			Fear (T3)
	ΔF	ΔR²	β	ΔF	ΔR²	β	ΔF	ΔR²	β	ΔF	ΔR²	β
Step 1	3.34*	.08*		2.12	.03		1.11	.02		.76	.05
Child sex			−.34*			.13			−.08			−.20
Child ethnicity			−.17			.15			−.12			.07
Step 2: T1 C emotion	9.41**	.15**	.72**	12.52**	.19**	.75**	7.34**	.11**	.50**	8.10**	.13**	.60**
Step 3: T3 M emotion	7.54**	.12**	.64**	10.04**	.14**	.69**	6.69**	.08**	.47**	7.42**	.12**	.61**
Step 4: HAI (T2)	3.56*	.08*	−.37*	4.52*	.10**	.46**	1.11	.02	.10	.84	.01	.07
Step 5	3.27*	.05*		4.05*	.05*		3.38*	.05*		3.56*	.05*
EE–Happy (T2)			.29*			−.19			−.32*			−.18
EE–Angry (T2)			.08			.37**			−.07			−.31*
EE–Sad (T2)			−.32*			.17			.38**			.21
EE–Fear (T2)			.11			−.36*			.25			.50**
Step 6	4.02*	.05*		3.28*	.05*		.96	.01		2.17	.02
HAI × EE–T1 Happy			.30*			−.30*			.15			−.14
HAI × EE–T1 Angry			.28*			.41**			.20			.10
HAI × EE–T1 Sad			.27			.23			.17			.20
HAI × EE–T1 Fear			.06			.18			.25			.11

Note. Table summarizes four separate regressions with T3 emotional expression. Sex coded 1 = girls, 2 = boys; Ethnicity coded 1 = African American, 2 = European American; C Emotion = child emotion; M emotion = mother emotion; HAI = Hostile Attributions of Intent; EE = Emotional Expectancies.

p < .05. **p < .01.

Although no specific hypotheses were advanced regarding interaction effects, for exploratory purposes the regression equations included interaction terms between adolescents’ hostile attributions and emotional expectancies. Analyses revealed that the two-way interaction between hostile attributions and emotional expectancies of happiness, and the two-way interaction between hostile attributions and emotional expectancies of anger, were uniquely related to adolescents’ expression of both happiness and anger. Follow-up analyses revealed that the negative association between hostile attributions and adolescent expression of happiness was stronger among adolescents with lower emotional expectancies of happiness, b = −.19, SE = .08, p < .001, as compared with adolescents with higher emotional expectancies of happiness, b = −.14, SE = .01, p < .05 (see Figure 1). Analysis of regions of significance of the data yielded both a lower and a higher bound of significance within the observed range of adolescent hostile attributions. More specifically, the slope between adolescents’ emotional expectancies of happiness and adolescent emotional expression of happiness with mother proved significant when hostile attributions was lower than 9.88, representing 6.7% of the sample, and higher than 19.08, representing 3.2% of the sample (i.e., shaded areas in Figure 1).

Figure 1.

The interaction of hostile attributions and emotional expectancies of happiness in predicting adolescents’ expression of happiness with their mother.

Among adolescents with higher emotional expectancies of anger, there was a significant negative association between hostile attributions and expression of happiness, b = −.14, SE = .07, p < .01, whereas for adolescents with lower emotional self expectancies of anger, there was no significant association between hostile attributions and expression of happiness, b = −.06, SE = .04, ns (see Figure 2). Analysis of regions of significance of the data yielded only a lower bound of significance within the observed range of adolescent hostile attributions. That is, the slope between adolescents’ emotional expectancies of anger and adolescent emotional expression of happiness proved significant when adolescent hostile attributions was lower than 10.86 (i.e., shaded areas in Figure 2), representing 8.5% of the sample, with no significant differences emerging above this value.

Figure 2.

The interaction of hostile attributions and emotional expectancies of anger in predicting adolescents’ expression of happiness with their mother.

In the equation for anger, follow-up analyses revealed that hostile attributions was significantly positively associated with expressions of anger among adolescents with low emotional expectancies of happiness, b = .16, SE = .07, p < .01, but was not significantly related to expression of anger among adolescents with high emotional expectancies of happiness, b = .04, SE = .04, ns (see Figure 3). Analysis of regions of significance of the data yielded only an upper bound of significance within the observed range of adolescent hostile attributions. Specifically, the slope between adolescents’ emotional expectancies of happiness and adolescent emotional expression of anger proved significant when adolescent hostile attributions was higher than 16.54 (i.e., shaded areas in Figure 3), representing 7.4% of the sample, with no significant differences emerging below this value.

Figure 3.

The interaction of hostile attributions and emotional expectancies of happiness in predicting adolescents’ expression of anger with their mother.

Among adolescents with high emotional expectancies of anger, the positive association between hostile attributions and adolescent anger was stronger, b = .17, SE = .08, p < .001, as compared with adolescents with low emotional expectancies of anger, b = .09, SE = .06, p < .05 (see Figure 4). Analysis of regions of significance of the data yielded only an upper bound of significance within the observed range of adolescent hostile attributions. Specifically, the slope between adolescents’ emotional expectancies of anger and adolescent emotional expression of anger proved significant when adolescent hostile attributions was higher than 18.24 (i.e., shaded areas in Figure 4), representing 4.8% of the sample, with no significant differences emerging below this value.

Figure 4.

The interaction of hostile attributions and emotional expectancies of anger in predicting adolescents’ expression of anger with their mother.

Predicting mother expression of emotion

Although no hypotheses were formed regarding predictors of mothers’ expression of emotion with adolescents, for the sake of thoroughness similar regression equations were computed with mothers’ emotions at T3 as the outcomes. In these equations, Step 2 contained the variable that represented mothers’ expression of emotion at T1 that matched the T3 emotion variable being predicted (e.g., in the regression predicting T3 happiness, Step 2 contained mothers’ happiness at T1; in the regression predicting T3 anger, Step 2 contained mothers’ anger at T1), whereas Step 3 contained the variable that represented adolescents’ expression of emotion at T3 that matched mothers’ emotion variable being predicted (e.g., in the regression predicting mother T3 happiness, Step 3 contained the adolescent’s happiness at T3; in the regression predicting mother T3 anger, Step 3 contained the adolescent’s anger at T3). Again, given space considerations, results from Steps 1, 2, and 3 are not described in the text but are included in the tables.

As shown in Table 4, adolescents’ hostile attributions predicted mothers’ expression of anger and fear, but were unrelated to mothers’ expression of happiness and anger. In addition, adolescents’ emotional expectancies predicted mothers’ expression of happiness and anger, but were unrelated to mothers’ expression of sadness or fear. Specifically, adolescents’ emotional expectancies of happiness was positively associated with mothers’ expressions of happiness and emotional expectancies of anger was negatively associated with mothers’ expressions of happiness, whereas adolescents’ emotional expectancies of happiness was negatively related to mothers’ expression of anger and emotional expectancies of anger was positively associated with mothers’ expression of anger. There were no significant two-way interactions between adolescent hostile attributions and emotional expectancies in predicting mother’s emotion.

Table 4.

Adolescents’ Hostile Attributions and Emotional Expectancies as Predictors of Mothers’ Expression of Emotion With Adolescent at T3.

	Happy (T3)			Angry (T3)			Sad (T3)			Fear (T3)
	ΔF	ΔR²	β	ΔF	ΔR²	β	ΔF	ΔR²	β	ΔF	ΔR²	β
Step 1	1.51	.01		3.21*	.07*		2.42*	.05*		1.14	.02
Child sex			−.12			.35*			−.15			.05
Child ethnicity			.18			.15			.31*			.17
Step 2: T1 M Emotion	11.71**	.17**	.60**	9.37**	.13**	.52**	7.45**	.09**	.43**	6.16**	.08**	.40**
Step 3: T3 C Emotion	10.24**	.14**	.52**	11.37**	.16**	.58**	5.88*	.07*	.35*	5.02*	.07*	.33*
Step 4: HAI (T2)	1.21	.04	.14	3.55*	.08*	.41*	1.26	.02	.19	3.10*	.07*	.35**
Step 5	3.65*	.05*		2.44	.04		1.95	.02		1.56	.02
EE–Happy (T2)			.30*			−.25			−.25			−.17
EE–Angry (T2)			−.28*			.20			.13			.15
EE–Sad (T2)			−.22			.10			.24			−.26
EE–Fear (T2)			.13			.18			.14			−.16
Step 6	1.10	.02		2.31	.04		1.88	.01		1.28	.03
HAI × EE–T1 Happy			.14			.30			.12			−.10
HAI × EE–T1 Angry			.08			11			.05			.06
HAI × EE–T1 Sad			.13			.16			.20			.23
HAI × EE–T1 Fear			.09			.23			.13			−.21

Note. Table summarizes four separate regressions with T3 emotional expression. Sex coded 1 = girls, 2 = boys; Ethnicity coded 1 = African American, 2 = European American; M Emotion = mother emotion; C Emotion = child emotion; HAI = Hostile Attributions of Intent; EE = Emotional Expectancies.

p < .05. **p < .01.

Discussion

The findings of this study represent the first empirical support, to the best of our knowledge, for theoretical arguments concerning the importance of distinguishing between attributions of intent and emotional expectancies as social information processing components that play a role in children’s expression of emotion (Teasdale, 2005; Weiner & Graham, 1984). The short-term longitudinal nature of the study makes the findings particularly noteworthy. Analyses revealed that individual differences in adolescents’ hostile attributions and emotional expectancies uniquely predicted expressions of emotions with their mother 6 months later. This pattern of findings supports arguments from social information processing theory (Dodge, 2006) that children’s thinking about emotions plays an important role in the origins of socioemotional behavior. Specifically, the results indicate that adolescents do indeed use both affective and social-cognitive information to interpret situational cues and guide their emotional responses However, caution should be exercised in interpreting these findings given that the magnitude of the effects of adolescent’s hostile attributions and emotional expectancies was not large in comparison with the effects accounted for by children’s earlier level of expression of emotion and for mothers’ expression of the same emotion in the immediate context. On the other hand, the fact that hostile attributions and emotional expectancies continued to account for a significant portion of the variance suggests that emotional expressiveness did not simply represent a stable tendency on the part of the adolescent to express a particular emotion or a tendency for adolescents to match the emotions expressed by their mother. Rather, children’s perceptions about their mother and their own expectancies about the emotions they would feel appear to play an independent role in shaping the emotions they express with their mother.

Adolescents’ Hostile Attributions and Expression of Emotion With Mother

The findings of the present study are consistent with previous evidence linking hostile attributions to negative patterns of mother-child interaction (Grace, Kelley, & McCain, 1993; MacKinnon-Lewis, Lamb, et al., 2001). The findings extend previous work by identifying associations between hostile attributions and particular emotions expressed by adolescents in the presence of their mother. Specifically, hostile attributions appear to contribute to more anger and less happiness expressed by adolescents when interacting with their mothers. This directly supports the proposition of attribution theory (Dodge, 2006; Weiner & Graham, 1984) that cognitive biases of hostility carried into new social situations can contribute to unfavorable patterns of social interaction, even in the absence of evidence to support hostile intent on the part of one’s social partner. Thus, the present findings suggest that hostile attributions may be one process that accounts for adolescent’s subjective experience of emotions. That is, to the extent that some adolescents are inclined to make hostile attributions of their mother’s behavior and others do not, similar situations can result in very different emotional consequences. Although we did not assess the actual content of the conversation between mother and child, and thus cannot say with certainty that the adolescent’s display of emotion was not the result of the topic being discussed, the findings do suggest that adolescent’s hostile attributions played at least a partial role in the level of happiness and anger youth expressed when interacting with their mother.

Adolescents’ Emotional Expectancies and Expression of Emotion With Mother

The pattern of associations observed between emotional expectancies and adolescents’ expression of emotion indicate that teens were fairly consistent in their expectations regarding the emotions that they would experience in hypothetical situations and in the emotions that they displayed when interacting with their mother. These findings are compatible with hypotheses outlined by Interaction Adaptation Theory (Burgoon, Stern, & Dillman, 2007) that emotional expectancies may be a form of emotional memory containing information about previous emotional experiences that activates emotions in new situations. The observed link between adolescents’ emotional expectancies and expression of emotion is significant for understanding how cognitive and emotional processes may operate conjointly to influence mother-adolescent relationship quality. Rather than being strictly online, in-the-moment, physiological-based reactions to environmental input, there appears to be a cognitive component to adolescents’ expression of emotion. This proposition is consistent with theorizing that cognitions and emotions may work together, instead of being separate subsystems of personality dynamics, to shape behavior at both the individual and dyadic level (Lemerise & Arsenio, 2000). Understanding such processes may lead to new insights regarding adolescent emotion regulation and social adaptation.

The data also point to interesting differential connections between particular emotional expectancies and discrete emotions displayed by adolescents. These connections between discrete emotional expectancies and nonparallel emotions suggest that adolescents’ cognitive appraisals of their own likely emotional state may serve as a protective or risk factor for subsequent emotional responses. That is, a cognitive-emotional model of oneself as having a tendency to react to situations with a positively valenced mood may operate, to some extent, as a self-fulfilling prophecy regardless of particular circumstances. Thus, youth with emotional expectancies that are positively valenced may be more likely to demonstrate relationship enhancing patterns of expression of emotion. In contrast, an ongoing emotional expectancy reflective of a proneness to anger can contribute to maladaptive emotion-cognition connections, lead to a faulty appraisal of a new event, and result in the activation of emotions that hinder optimal social interactions.

Interaction Between Adolescents’ Hostile Attributions and Emotional Expectancies in Predicting Expression of Emotion With Mother

The present study also offers evidence to support theories to explain how attributions of intent (Brody et al., 1996; Crick & Dodge, 1994) and emotional expectancies (Arsenio & Lemerise, 2004; Izard et al., 2002) operate in conjunction to predict emotional expressions in social situations. Emotional expectancies of happiness seemed to have a protective function for more positive and less negative emotional expression with mother among children with hostile attributions. That is, adolescents who attributed hostile intent to their mother’s behavior, but expected to be happy, were more likely to express happiness and less likely to express anger with their mother than adolescents who attributed hostile intent to their mother and expected to be less happy and more angry. In particular, hostile attributions were significantly related to less expression of happiness with mother among adolescents with low emotional expectancies of happiness, but were not related to the expression of happiness among adolescents with high emotional expectancies of happiness. Similarly, hostile attributions were significantly related to higher expression of anger among adolescents with low emotional expectancies of happiness, but were not related to the expression of anger among adolescents with high emotional expectancies of happiness. Indeed, the adolescents who expressed the most anger with their mothers were those who not only made hostile attributions but also had higher emotional expectancies of anger. This pattern of findings points to the complex ways in which attributions of self and other may combine to influence adolescents’ emotions. However, caution should be exercised in interpreting the interaction effects given the post hoc nature of such analyses and the fact that regions of significance tests indicate that effects were accounted for by extreme scores of only 3.2% to 8.5% of the sample.

Adolescent Hostile Attributions and Emotional Expectancies and Mother Expression of Emotion

Adolescents’ hostile attributions also predicted mothers’ expression of emotion during their interaction together. Specifically, youth who attributed hostile intent to their mother’s behavior had mothers who expressed more anger and more fear. Youth’s emotional expectancies were also found to be associated with mothers’ expression of discrete emotions. For example, youth who expected to react with happiness had mothers who expressed more happiness and less anger, whereas youth who expected to react with anger had mothers who expressed more anger and less happiness. One interpretation of these findings is that youth hostile attributions and emotional expectancies may be based on a history of mother-child interaction. To the extent that mothers manifest a similar pattern in expressing particular emotions with their children over time, children may incorporate knowledge of their own reactions to mothers’ emotions into their self-expectancies. Alternatively, it may be that mother’s emotions are influenced by the adolescent’s own expressions of emotions that are themselves guided by the teen’s hostile attributions and emotional expectancies. Because the data of this study is correlational in nature, it is impossible to determine the direction of effect in the observed connections.

Adolescent Gender and Ethnicity

Although the pattern of associations between adolescent attributions, emotional expectancies, and the emotions expressed by adolescent and mother held true after controlling for child gender and ethnicity, it of interest to note that there were several mean differences in scores based on child gender and ethnicity. Consistent with previous findings, boys held more hostile attributions of mothers’ behavior than girls (Grace et al., 1993), and African American adolescents had higher hostile attributions than European American adolescents (MacKinnon-Lewis, Lamb, et al., 2001). With regard to emotional expectancies, girls had higher expectancies of happiness and sadness, whereas boys had higher expectancies of anger. In the expression of emotion with mothers, girls displayed more happiness than boys, and European American adolescents expressed more happiness than African American adolescents. With regard to mothers’ expression of emotion, mothers of girls expressed more sadness, whereas mothers of boys expressed more anger. These findings suggest that gender stereotypes and cultural differences may play a role in the content of adolescents’ information processing knowledge and patterns of emotions expressed in mother-adolescent dyads and should be more fully explored in future research.

Limitations of Study

There are a number of other limitations to the study that should be kept in mind when interpreting the data. For instance, the overall rates of adolescents’ and mothers’ expression of negative emotions, particularly sadness and fear, were quite low. In one sense, this is to be expected given evidence that both teens and parents report their relationship to be generally positive (Larson et al., 1996; Shearer et al., 2005). At the same time, however, the contrived nature of the interaction session and the fact that the mother-adolescent dyads were aware they were being observed might have limited the natural expressions of emotions. Additional research in more natural settings or using multiple contexts of parent-child interaction might provide more accurate data regarding the emotions expressed between mothers and youth.

Another limitation is that the study only assessed adolescents’ expectancies regarding their own emotions. This decision was based in part on a belief that, although adolescents may have a clear understanding of another person’s likely emotional experiences, it is their own emotional expectancies that are most salient to their behavior. However, it may be that expectations regarding the emotions that an interactive partner is likely to experience are equally important in adolescents’ decision making with regard to what behaviors to enact. Future research should consider both self and other emotional expectancies to gain a clearer understanding of the sources of influence on adolescents’ expression of emotion.

The fact that adolescents’ hostile attributions and emotional expectancies were based on responses to hypothetical, not real, situations also restricts the interpretations that can be made of the findings. Participants’ responses in real situations might differ from those they gave to the hypothetical vignettes. In addition, although this research describes intriguing longitudinal associations between adolescents’ emotional expectancies regarding mothers’ hypothetical behavior and the emotions that they express when interacting with their mother, the data do not provide any information about causal relations among the variables. Experimental studies that implement mood induction techniques in adolescents, followed by assessments of emotional expectancies, may prove useful in identifying more specific causal pathways between emotional expectancies and expression of emotion.

In spite of these limitations, the results provide new evidence on the ways that adolescents’ hostile attributions and emotional expectancies relate to patterns of expression of emotion between mothers and teens. The results suggest that it would be valuable to expand the focus on emotion in the cognitive-contextual framework and, in particular, to explore affective processes involved in adolescents’ construction of self-knowledge. As family expression of emotion is hypothesized to be a significant contributor to early adolescent’s self-regulatory capabilities (Conger & Ge, 1999; Magai & McFadden, 1995), investigating the role of children’s emotional expectancies in their understanding and expression of emotions could provide further insight into the processes involved in adolescent’s risk for maladjustment.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by William T. Grant Foundation Grant 017396218.

Author Biographies

Eric W. Lindsey is a professor of applied psychology at Penn State University Berks Campus, Reading, Pennsylvania. He received his MS and PhD degrees in human development and family studies from Auburn University. His research interests focus connections between parent-child relationships and children’s social competence with peers.

Carol MacKinnon-Lewis is a professor in the Department of Child and Family Studies, at the University of South Florida. She received her MS and PhD degrees in child development from the University of Georgia. Her research interests include adolescents’ social information processing and youth aggression.

James M. Frabutt is a faculty member in the Mary Ann Remick Leadership Program in the Alliance for Catholic Education, at Notre Dame University. He received his MS and PhD in human development and family studies from the University of North Carolina at Greensboro.

Jessica Campbell Chambers is health science administrator at the Division of Clinical Neuroscience, Development and Behavior Research of the National Institute of Drug Abuse. She received her MS and PhD degrees in human development and family studies from the University of North Carolina at Greensboro.

References

Aiken

L. S.

West

S. G.

(1991). Multiple regression: Testing and interpreting interactions. New York, NY: Sage.

Arsenio

W. F.

Gold

Adams

(2004). Adolescents’ emotion expectancies regarding aggressive and nonaggressive events: Connections with behavior problems. Journal of Experimental Child Psychology, 89, 338-355.

Arsenio

W. F.

Lemerise

(2004). Aggression and moral development: Integrating the social information processing and moral domain models. Child Development, 75, 987-1002.

Brody

G. H.

Arias

Fincham

F. D.

(1996). Linking marital and child attributions to family processes and parent-child relationships. Journal of Family Psychology, 10, 408-421.

Bugental

D. B.

Johnston

(2000). Parental and child cognitions in the context of the family. Annual Review of Psychology, 51, 315-344.

Burgoon

J. K.

Stern

L. A.

Dillman

(2007). Interpersonal adaptation: Dyadic interaction patterns. New York, NY: Cambridge University Press.

Burks

V. S.

Laird

R. D.

Dodge

K. A.

Pettit

G. S.

Bates

J. E.

(1999). Knowledge structures, social information processing, and children’s aggressive behavior. Social Development, 8, 220-235.

Chung

G. H.

Flook

Fuligni

A. J.

(2009). Daily family conflict and emotional distress among adolescents from Latin America, Asian, and European backgrounds. Developmental Psychology, 45, 1406-1415.

Cohen

(1975). Applied multiple regression/correlation analysis for the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum.

10.

Collins

Laursen

(2006). Parent-adolescent relationships. In Noller

Feeney

J. A.

(Eds.), Close relationships: Functions, forms and processes (pp. 111-125). New York, NY: Psychology Press.

11.

Conger

R. D.

Conger

K. J.

Elder

G. H.

Jr. Lorenz

F. O.

Simons

F. L.

Whitebeck

L. B.

(1993). Family economic stress and adjustment of early adolescent girls. Developmental Psychology, 29, 206-219.

12.

Conger

R. D.

(1999). Conflict and cohesion in parent adolescent relations: Changes in emotional expression from early to mid-adolescence. In Cox

M. J.

Brooks-Gunn

(Eds.), Conflict and cohesion in families: Causes and consequences (pp. 185-206). Mahwah, NJ: Lawrence Erlbaum.

13.

Conger

R. D.

Lorenz

F. O.

Elder

G. H.

Jr. Melby

J. N.

Simons

R. L.

Conger

K. J.

(1991). A process model of family economic pressure and early adolescent alcohol use. Journal of Early Adolescence, 11, 430-449.

14.

Creswell

Murray

L. M.

Cooper

P. J.

(2010). Intergenerational transmission of anxious information processing bias. In Hadwin

J. A.

Field

A. P.

(Eds.), Information processing biases in child and adolescent anxiety (pp. 279-295). Chichester, UK: Wiley-Blackwell.

15.

Crick

Dodge

K. A.

(1994). A review and reformulation of social information-processing mechanisms of children’s social adjustment. Psychological Bulletin, 115, 74-101.

16.

Dodge

K. A.

(2006). Translational science in action: Hostile attributional style and the development of aggressive behavior problems. Development and Psychopathology, 18, 791-814.

17.

Grace

N. C.

Kelley

M. L.

McCain

A. P.

(1993). Attribution processes in mother-adolescent conflict. Journal of Abnormal Child Psychology, 21, 199-211.

18.

Harris

P. L.

(1985). What children know about the situations that provoke emotion? In Lewis

Saarni

(Eds.,) The socialization of emotions (pp. 161-185). New Mexico, US: Springer.

19.

Izard

C. E.

(1977). Human emotions. New York, NY: Plenum Press.

20.

Izard

C. E.

(2001). Emotional intelligence or adaptive emotions? Emotion, 1, 249-257.

21.

Izard

C. E.

Fine

Mostow

Trentacosta

Campbell

(2002). Emotion processes in normal and abnormal development and preventive intervention. Development and Psychopathology, 14, 761-787.

22.

Krettenauer

Jia

Mosleh

(2011). The role of emotion expectancies in adolescents’ moral decision making. Journal of Experimental Child Psychology, 108, 358-370.

23.

Larson

R. W.

Richards

M. H.

Moneta

Holmbeck

Duckett

(1996). Changes in adolescents’ daily interactions with their families from ages 10 to 18: Disengagement and transformation. Developmental Psychology, 32, 744-754.

24.

Lemerise

E. A.

Arsenio

W. F.

(2000). An integrated model of emotion processes and cognition in social information processing. Child Development, 68, 1181-1197.

25.

Lewis

M. D.

(2000). Emotional self-organization at three time scales. In Lewis

M. D.

Granic

(Eds.), Emotion development, and self-organization: Dynamic systems approaches to emotional development (pp. 37-69). New York, NY: Cambridge University Press.

26.

MacKinnon-Lewis

C. E.

Castellino

D. R.

Brody

G. H.

Fincham

F. D.

(2001). A longitudinal examination of the associations between fathers’ and children’s attributions and negative interactions. Social Development, 10, 473-487.

27.

MacKinnon-Lewis

C. E.

Lamb

M. E.

Arbuckles

Baradaran

L. P.

Volling

B. L.

(1992). The relationship between biased maternal and filial attributions and the aggressiveness of their interactions. Development and Psychopathology, 4, 403-415.

28.

MacKinnon-Lewis

C. E.

Lamb

M. E.

Hattie

Baradaran

L. P.

(2001). A longitudinal examination of the associations between mothers’ and sons’ attributions and their aggression. Development and Psychopathology, 13, 69-81.

29.

MacKinnon-Lewis

C. E.

Volling

B. L.

Lamb

M. E.

Dechman

Rabiner

Curtner

M. E.

(1994). A cross-contextual analysis of boys’ social competence: From family to school. Developmental Psychology, 30, 325-333.

30.

Magai

McFadden

S. H.

(1995). The role of emotions in social and personality development: History, theory, and research. New York, NY: Plenum Press.

31.

Preacher

K. J.

Curran

P. J.

Bauer

D. J.

(2006). Computational tools for probing interactions in multiple linear regression, multilevel modeling, and latent curve analysis. Journal of Educational and Behavioral Statistics, 31, 437-448.

32.

Reijntjes

Thomaes

Kamphuis

J. H.

Bushman

B. J.

De Castro

B. O.

Telch

M. J.

(2011). Explaining the paradoxical rejection-aggression link: The mediating effects of hostile intent attributions, anger, and decreases in state self-esteem on peer rejection-induced aggression in youth. Personality and Social Psychology Bulletin, 37, 955-963.

33.

Shanahan

McHale

S. M.

Osgood

D. W.

Crouter

A. C.

(2007). Conflict frequency with mothers and fathers from middle childhood to late adolescence: Within- and between-families comparisons. Developmental Psychology, 43, 539-550.

34.

Shearer

C. L.

Crouter

A. C.

McHale

S. M.

(2005). Parents’ perceptions of changes in mother-child and father-child relationships during adolescence. Journal of Adolescent Research, 20, 662-684.

35.

Teasdale

J. D.

(2005). Multi-level theories of cognition-emotion relations. In Dalgleish

Power

M. J.

(Eds.), Handbook of cognition and emotion (pp. 665-682). Chichester, UK: John Wiley.

36.

Weiner

Graham

(1984). An attributional approach to emotional development. In Izard

Kagan

Zajonc

(Eds.), Emotion, cognition and behavior (pp. 167-191). Cambridge, MA: Cambridge University Press.

37.

Yeager

D. S.

Miu

Powers

Dweck

C. S.

(2013). Implicit theories of personality and attributions of hostile intent: A meta-analysis, an experiment, and a longitudinal intervention. Child Development, 84, 970-988.