هدیه ای ناامید کننده: مدیران جهت مند و موقعیتی ابراز هیجانی

Abstract Inferences about emotions in children are limited by studies that rely on only one research method. Convergence across methods provides a stronger basis for inference by identifying method variance. This multimethod study of 116 children (mean age = 8.21 years) examined emotional displays during social exchange. Each child received a desirable gift and later an undesirable gift after performing tasks, with or without mother present. Children’s reactions were observed and coded. Children displayed more positive affect with mother present than with mother absent. Independent ratings of children by adults revealed that children lower in the personality dimension of Agreeableness displayed more negative emotion than their peers following the receipt of an undesirable gift. A curvilinear interaction between Agreeableness and mother condition predicted negative affect displays. Emotional assessment is discussed in terms of links to social exchange and the development of expressive behavior.

Introduction The study of emotion in children is complex and multifaceted. Conceptual ambiguity underlies the construct of emotion in part because it refers to potential for behavior that may or may not translate into overt behavior. Like other predispositions, emotions require specific eliciting conditions for the potential to be activated. To understand emotional processes, and by extension emotional assessment, it is necessary to specify contextual variables that activate emotion-related structures. Specification of these conditions goes hand-in-glove with the nature of the emotion in question (Cole, Martin, & Dennis, 2004). For example, if the focal process is negative emotion, then frustrating conditions will probably be more effective elicitors than will reward conditions. This characterization may apply better to emotional traits than to emotional states. That is, a predisposition to negative emotionality or regulation of negative emotion may require specification of multiple comparative eliciting conditions to a greater extent than would a single negative emotional display. Even here, however, conditions that could and could not elicit the display would be of interest to researchers. Taken together, this line of reasoning implies that experiments may be a prime tool for assessing emotional processes in children. Experiments are not commonly regarded as tools in the assessment arsenal, but these procedures allow children to be randomly assigned to conditions in which contextual variables are manipulated and not merely correlated. If the hypotheses are correct, then variations in conditions will elicit emotional relations relative to another control condition. Experiments are the most powerful method when causal inference is the main goal of research (e.g., Shadish et al., 2002 and West et al., 2000). Nevertheless, experiments are not without their critics. Perhaps the most common criticism is that outcomes of laboratory experiments lack external validity. That is, they are constrained, artificial, and unlikely to generalize beyond the confines of the laboratory walls, if even that far. Recently, theorists have noted that external validity is more important than previously recognized because outcomes that do not generalize have implications for larger questions of boundary conditions and construct validity (e.g., Shadish and Cook, 2009, Shadish et al., 2002 and West and Graziano, in press). One step toward a resolution involves the recognition of the need for convergence across multiple methodologies. Every methodology contains potential limitations, so outcomes that appear across methodologies imply that the outcome transcends method variance. In the case of emotional assessment in children, this logic implies that experimental paradigms may be especially valuable for manipulating eliciting conditions for emotional displays in children and making stronger causal inference (Cole et al., 2004). When experimental methods are combined with observational and correlation methodologies, however, alternative explanations centered on constraining measures, artificiality, and lack of generality outside the laboratory are weakened, if not largely eliminated, as plausible explanations (Zeman, Klimes-Dougan, Cassano, & Adrian, 2007). The research presented here used a converging multimethod study to probe hypotheses about emotional assessment in children. In particular, we used experimental and correlation procedures, combined with observational coding and reports of mothers and knowledgeable adult informants, to examine emotional displays in children. Substantively, this research focused on emotional displays and their variation as a function of eliciting conditions. Presumably, if a child exhibits one display in Setting A but a different one in Setting B, the difference could be a manifestation of emotion regulation. It could also be the result of several other processes such as differential sensitivity to the eliciting conditions. To probe all of these processes, it is important to have a working definition of emotion regulation. Thompson (1994) observed that implicit notions of emotion regulation are so powerful that researchers often do not provide a clear explicit definition of the phenomenon. He then offered this working definition: “Emotion regulation consists of the extrinsic and intrinsic processes responsible for monitoring, evaluating, and modifying emotional reactions, especially their intensive and temporal features, to accomplish their goals” (pp. 27–28). Ultimately, Thompson suggested that emotion regulation is not easily defined because it refers to a range of dynamic processes, “each of which may have its own catalysts and control processes” (p. 52). Subsequent analysis by Cole and colleagues (2004) came to similar conclusions about the elusiveness of a consensus definition. Until such consensus is reached, emotion regulation researchers must provide working definitions through discussion of their key constructs. This research focused on the assessment of emotion and emotion regulation within the context of social exchanges because they represent an opportunity for multimethod research. They can be observed naturalistically, described by third parties, and manipulated experimentally. Perhaps more important, social exchange is recognized as a universal forum in which emotion regulation is required (Laursen & Graziano, 2002). Anthropologists identify social exchange as a panhuman experience, noting that children in every known culture must learn rules for the giving and receiving of gifts (Harris, 1968). Developing social exchange skills (and the related emotion regulation skills) is an important component of forming alliances and building coalitions. Thus, children must learn culturally appropriate ways to regulate their emotional reactions during social exchange. The current study brought this panhuman experience into the laboratory for a more focused analysis. Anthropologists and social exchange theorists note that social exchange outcomes, and perhaps even processes, are “conditioned” by the nature of the persons participating in the exchange (e.g., Harris, 1968). How relationship contexts condition exchanges has been a topic of empirical investigation but limited experimental research. Zeman and Garber (1996) asked first, third, and fifth graders to respond to vignettes in terms of displays. Questions included forecasting whether they would show that they were mad, sad, or feeling pain. All children reported that mothers and fathers would be more understanding and accepting of emotional displays than would peers. Children reported that the primary reason for controlling their emotional expressions was the expectations of negative interpersonal interactions following displays. These conclusions are certainly plausible. In a large three-study multimethod program of research, Tobin, Graziano, Vanman, and Tassinary (2000) asked university students to report to the laboratory with a friend. Students were randomly assigned to describe emotion-provoking slides either to their friend or to a stranger (another student’s friend) with whom they believed they were communicating through a video camera that was covertly recording their emotional reactions. They also rated how much emotion the slides evoked and how much effort they exerted to control their emotions. Sex and the personality dimension of Agreeableness were significant predictors of both emotional experience and efforts to control emotion. When participants reported experiencing intense negative emotions, observers blind to the experimental conditions and participants’ personalities reported that participants also appeared to be exerting greater efforts to control emotions. Participants also forecasted that they would communicate greater emotional intensity to friends than to strangers, but in actual emotional communication there was no evidence of a difference. Despite their plausibility, vignette responses might not forecast actual emotional display behavior well. Different methods may lead to somewhat different conclusions, but it is reasonable to expect the presence of adults, and mothers in particular, to influence children’s affect displays during social exchange. When gifts are desirable, adults can stand witness to positive emotion displays and their ability to support and maintain social bonds. When gifts are undesirable, however, adult presence may serve a different function. In this case, an undesirable gift would likely be a source of frustration that is difficult for a child to control. If the child were to generate a high-amplitude display of negative emotion, it could be disruptive to social relations. In this sense, then, adults serve double duty: They support positive displays in the presence of a desirable gift and inhibit relationship-damaging displays of negative affect in response to an undesirable gift. Empirical research generally supports this proposition about the influence of onlookers (e.g., Cole, 1986, Shipman et al., 2001 and Zeman and Garber, 1996). Moving from the level of the group to the level of the individual, researchers have identified systematic patterns of individual differences in emotion regulation. Specifically, developmental theorists suggested a connection between two major personality dimensions and regulation: Ahadi and Rothbart (1994) linked Agreeableness and Conscientiousness to the temperamental process of “effortful control,” a form of executive regulatory temperament connected to the deployment of attention. Effortful control processes allow the child to direct attention away from blocked goals and toward adaptive action. These processes are flexible and can be centered on interactions with people in the case of Agreeableness or on interaction with tasks in the case of Conscientiousness. Rothbart and colleagues theorized that Agreeableness may emerge developmentally from the ability to shift and focus attention and/or inhibit action in emotionally evocative interpersonal situations (e.g., Ahadi and Rothbart, 1994 and Rothbart and Bates, 1998). Evidence supports the theoretical link between effortful control and Agreeableness (Cumberland-Li et al., 2004 and Jensen-Campbell and Graziano, 2005). These results suggest that individuals higher in Agreeableness regulate their emotions more than their peers to maintain smooth interpersonal relations. Consistent with these findings, Tobin and colleagues (2000) found that participants high in Agreeableness reported greater emotional experience as well as greater efforts to control negative emotions. Agreeableness is connected to motives to maintain smooth interpersonal relations (Graziano and Eisenberg, 1997, Graziano and Habashi, 2010 and Graziano and Tobin, 2009). This individual difference is pervasive in social cognition (Graziano, 1994) and may be the most socialized of the personality dimensions (Bergeman et al., 1993 and Kohnstamm et al., 1998). It is not completely clear, however, how this personality dimension connects to emotional behavior during development. Given the underlying motive to maintain smooth interpersonal relations, it is expected that Agreeableness will be linked to children’s regulation of emotion in social situations (e.g., Jensen-Campbell & Graziano, 2005). As anthropologists have noted, other factors beyond the individual, such as interaction partners, influence emotional expression in individuals. Specifically, the presence of a socializing agent such as a mother influences a child’s emotional behavior (Cassano et al., 2007 and Fabes et al., 2002). In some situations, the presence of a socializing agent may place added pressure on the child to express the socially prescribed emotional behavior (Bugental and Goodnow, 1998, Halverson et al., 1994 and Tobin and Graziano, 2010). Thus, beyond individual differences, the presence of a mother also influences a child’s expressions of negative emotions. One possibility for processes underlying maternal presence may be explained by analogy with the cognitive developmental research on “production deficiencies” (Flavell, Green, & Flavell, 1995; for a critical discussion of these and related utilization deficiency mechanisms, see Waters, 2000). Younger children may perform less well than older children on memory tasks not from a lack of inherent memory capacity per se but rather from a failure to deploy strategic rehearsal strategies. When given focused instruction in metacognitive strategies, younger children may perform as well as older children. By analogy, children who appear to be low in regulatory skills may be like younger children; they need remedial focused training or to be “reminded.” The presence of a mother may serve as a reminder, particularly to children low in Agreeableness. Once reminded of the need to attend to norms for the display of negative emotions, children low in regulatory skills may perform in ways comparable to their more highly regulated (and presumably more internalized) peers. If this analogy is valid, then a major difference between children who regulate negative emotion when receiving an undesirable gift and those who do not is not in regulatory skills per se but rather in the social–cognitive salience of the need to regulate in this particular context. A mother may serve as a reminder of the need to regulate in the context of an undesirable gift for such children. Following this logic further, if maternal presence is primarily a sophisticated way to overcome production deficiencies in motivation, in effect to “remind” children of the need to regulate emotional displays in this context, and if some children will profit from the reminders more than other children, then maternal presence will have less systematic influence on children for whom the need to regulate negative affect is already salient. By this logic, children high in Agreeableness will be influenced less by the presence of their mothers because the need to regulate is already salient for them regardless of external reminders; they have internalized the socialization of emotion displays (for a related discussion, see Graziano, Habashi, Sheese, & Tobin, 2007, Study 4). One of the focal issues in this program of research is how individual differences in regulatory skills combine with situational and contextual factors in affecting regulatory processes and behavioral expression. In the current study, we examined these predictors in children between 5 and 10 years of age because this age range has been studied in this paradigm extensively (e.g., Cole, 1986 and Saarni, 1984) and children in this age range show considerable variability in emotional responding to gifts (e.g., Kieras et al., 2005, Saarni, 1989 and Underwood, 1997). In the current study, we investigated the relation between emotion regulation in a social situation and children’s Agreeableness using the reports of informed adult caregivers. In addition to individual differences, we also examined the presence or absence of a mother as a predictor of emotion regulation. We hypothesized that children high in Agreeableness would demonstrate greater emotion regulation than children low in Agreeableness. In other words, videotape coders blind to the participants’ Agreeableness scores would rate children high in Agreeableness as experiencing more positive or neutral emotion, and less negative emotion, after the receipt of an undesirable gift than that experienced by children low in Agreeableness. We also hypothesized that children high in Agreeableness would show greater emotion regulation than their peers regardless of the mothers’ presence. We expected children low in Agreeableness, however, to regulate their emotions more when a mother was present than when a mother was absent. That is, children low in Agreeableness need the presence of their mothers as a reminder of socially appropriate behavior when given an undesirable gift, unlike their high-Agreeableness peers, who presumably have already internalized the standard. As in past research (Cole, 1986, McDowell et al., 2000 and Saarni, 1984), we expected girls to show greater emotion regulation than boys.

Results The criterion variables were mean scores for positive and negative affect at the time each child received an undesirable gift. Following procedures described by Aiken and West (1991) for assessing interactions between categorical and continuous variables (pp. 116–136), we used regression procedures to analyze the Agreeableness × Mother Condition interactions. To facilitate interpretation of regression coefficients, we used dummy codes for mother condition. Displays of positive affect Based on the recommendations of Aiken and West (1991), we conducted centered cross-product regression analyses to test hypotheses. In keeping with their recommendations, mother condition, sex, age, and Agreeableness main effects were entered first, followed by the Agreeableness × Mother Condition cross-product term. The first model that included the main effects was significant, F(4, 111) = 4.30, p < .01, R2 = .13. Mother condition was a significant predictor of positive displays of affect, β = .25, t(111) = 2.81, p < .01. Children in the presence of their mothers (M = 1.67, SD = 0.75) displayed more positive affect than did children without their mothers present (M = 1.33, SD = 0.49). Sex also was a significant predictor of positive emotional display, β = .23, t(111) = 2.52, p < .01. Girls (M = 1.65, SD = 0.70) displayed significantly more positive emotion than did boys (M = 1.35, SD = 0.56) when receiving an undesirable gift. There was no evidence that age, Agreeableness, or the Agreeableness × Mother Condition interaction were significant predictors of positive affect displays, β = .10, t(111) = 1.15, p = .25; β = −.10, t(111) = −1.12, p = .27; and β = −.01, t(110) = −0.06, p = .95, respectively. Displays of negative affect A similar data analytic approach was used with displays of negative affect. Agreeableness was a significant predictor of negative affect displays following the receipt of an undesirable gift, β = −.22, t(111) = −2.34, p < .02. Children high in Agreeableness displayed less negative affect than did children low in Agreeableness. There was no evidence that age, sex, and mother condition were significant predictors of negative affect displays, β = −.00, t(111) = −0.05, p = .96; β = .03, t(111) = 0.30, p = .77; and β = −.06, t(111) = −0.59, p = .55, respectively. We also predicted a significant interaction between Agreeableness and mother condition. When the Agreeableness × Mother Condition interaction was entered into a regression analysis with Agreeableness, mother condition, and sex, there was no evidence that the interaction term predicted displays of negative affect when receiving an undesirable gift, β = .01, t(110) = 0.05, p = .96. Because the predicted linear interaction was not significant, we also explored other possible descriptions of the outcome. Thus, we also examined the interaction between Agreeableness and mother condition in a curvilinear regression analysis ( Aiken & West, 1991). That is, we examined the hypothesis that the presence of a mother has less influence on the emotional displays of children who are at the extremes of the Agreeableness dimension than it has for children at moderate levels of Agreeableness. In effect, we examined the possibility that children’s personalities contributed to the “situation” ( Halverson and Wampler, 1997 and Kelley et al., 2003). When the second-order Agreeableness term and the second-order Agreeableness × Mother Condition interaction were entered with Agreeableness, mother condition, sex, and the Agreeableness × Mother Condition interaction, the second-order Agreeableness × Mother Condition interaction was a significant predictor of negative displays of emotion, β = .31, t(109) = 2.22, p < .03. Fig. 1 shows the interaction effect between Agreeableness and mother condition on displays of negative emotion. Agreeableness, mother condition, and negative affect. Fig. 1. Agreeableness, mother condition, and negative affect. Figure options Based on the recommendations of Aiken and West (1991), we conducted follow-up analyzes of simple slopes for this significant curvilinear interaction. The mother condition effect was marginally significant for children in the middle range of Agreeableness, β = −.23, t(109) = −1.86, p < .07, but there was no evidence that the mother condition effect was significant for children 1 standard deviation above the mean on Agreeableness, β = .00, t(109) = 0.02, p = .98, or for children 1 standard deviation below the mean on Agreeableness, β = −.11, t(109) = −0.82, p = .42. Taken together, these findings suggest that the presence of a mother is most useful in promoting socially appropriate emotion displays in children of moderate Agreeableness. Children high and low in Agreeableness may be more “traited” for displays of negative emotion, and the presence of a mother does not alter their response to the disappointing gift. In the case of children high in Agreeableness, theory suggests that appropriate display rules have already been internalized. In contrast, children low in Agreeableness require more than the reminder or their mothers’ presence to override their natural tendency to respond negatively to the disappointing gift. These results lend support to the hypothesis that the Agreeableness × Mother Condition interaction is a significant predictor of emotion displays, at least for displays of negative affect. There was no evidence of a curvilinear Agreeableness × Mother Condition interaction effect in the prediction of positive affect, β = −.07, t(109) = −0.53, p = .60.