Parenting and Child Characteristics in the Prediction
by user
Comments
Transcript
Parenting and Child Characteristics in the Prediction
M e r r i l l - P a l m e r Q u a rt e r ly , V o l . 5 6 , N o . 4 Parenting and Child Characteristics in the Prediction of Shame in Early and Middle Childhood Rosemary S. L. Mills, Kimberley A. Arbeau, Debra I. K. Lall, and Amy E. De Jaeger University of Manitoba We examined individual differences in shame responding in early childhood and predictive relations with shame proneness in middle childhood. Child shame responding, parental shaming, and child temperamental inhibition were assessed at Time 1 (n = 225, aged 3–4 years), shame responding was reassessed at Time 2 (n = 199, aged 5–7 years), and shame proneness was assessed at Time 3 (n = 162, aged 7–9 years). Shame responding was assessed from emotionexpressive reactions to failure, parental shaming from self-reports and spouse reports, temperamental inhibition from mother and father ratings, and shame proneness from hypothetical scenarios. Girls showed more shame than boys by school age. Increased shame responding between preschool age and school age was predicted, for girls, from lower inhibition or higher mother shaming and, for boys, from higher mother shaming if boys were highly inhibited. Shame responding at preschool age predicted higher or lower shame proneness in middle childhood conditional on gender and parenting. Shame plays a central role in behavior. It is a key emotional barometer of a person’s felt worth or value that motivates and regulates thinking, feeling, and behavior related to preserving acceptance by others (Dickerson, Gruenewald, & Kemeny, 2004; Dickerson & Kemeny, 2004; H. B. Lewis, 1971; Rosemary S. L. Mills, Department of Family Social Sciences; Kimberley A. Arbeau, Department of Family Social Sciences; Debra I. K. Lall, Department of Psychology; Amy E. De Jaeger, Department of Psychology. This research was supported by Canadian Institutes of Health Research grant MOP-57670. Special thanks are extended to the children and parents who made this research possible and to research assistants Gorette Imm, Leanne Mak, Nadine Sigvaldason, Brooke Singer, Bobbi Walling, and Andrea Winther. Address correspondence to Rosemary Mills, Department of Family Social Sciences, University of Manitoba, Winnipeg, Canada R3T 2N2. Phone: (204) 474-9432. E-mail: rosemary_mills@ umanitoba.ca. Merrill-Palmer Quarterly, October 2010, Vol. 56, No. 4, pp. 500–528. Copyright © 2010 by Wayne State University Press, Detroit, MI 48201. 500 Shame in Early and Middle Childhood 501 M. Lewis, 1992; Tangney & Fischer, 1995). States of shame are painful and aversive, involving a sense of helplessness about the self, a desire to hide or disappear, difficulty sustaining social interaction, and even difficulty speaking fluently and thinking coherently (H. B. Lewis, 1971; M. Lewis, 1992; Weiner, 1986). As such, they are potent experiences that play a role in both adaptive and maladaptive development. Shame is essential for normal development, helping to motivate socially acceptable behavior and efforts to improve the self. When it becomes a dominant emotion, however, it can be maladaptive (Barrett, 1998; M. Lewis, 1992; Schore, 1996), as indicated by links between proneness to shame (a disposition to respond with shame) and a wide range of mental and physical health problems beginning at least by the school years (Dickerson et al., 2004; Ferguson, Stegge, Miller, & Olsen, 1999; Tangney, Burggraf, & Wagner, 1995). Despite the recognized developmental significance of shame, there has been a paucity of research on the sources of individual differences in shame responding in young children. The purpose of the present study was to address this gap in the literature by examining factors predicting shame responding in normally developing children in early childhood. Shame emerges by 3 years of age (Lagattuta & Thompson, 2007; M. Lewis, 1992, 2007). Attributions of responsibility result in specific emotions, with self-attributions for failure precipitating shame when the focus is global and guilt when the focus is specific (e.g., Tangney & Dearing, 2002). By age 3, children have the cognitive prerequisites for making global self-attributions. They have self-awareness, they are able to take the view of others and think about others’ views toward them, and they have acquired standards, rules, and goals and can evaluate themselves as succeeding or failing to meet them. They also have acquired the ability to evaluate themselves in terms of responsibility and, when accepting responsibility, they can focus on the global self or on specific performance as the cause. Also by age 3, children’s understanding of others has developed sufficiently that they are becoming more sensitive to others’ actual or imagined judgments and are more vulnerable to feelings of shame (Lagattuta & Thompson, 2007). Not surprisingly, then, individual differences in shame responding emerge by this age (e.g., Barrett, Zahn-Waxler, & Cole, 1993; Kelley, Brownell, & Campbell, 2000; M. Lewis, Alessandri, & Sullivan, 1992). Given the developmental significance of proneness to shame, a better understanding of the sources of early individual differences is needed. Thus, the present study focused on sources in early childhood. Theories of emotional development (Denham, 1998; Saarni, 1999) suggest that individual differences in the experience, expression, regulation, and understanding of emotions derive from the interactive effects of 502 Merrill-Palmer Quarterly factors both internal and external to the child. Given the crucial role parents have as socializers, parental socialization is likely to be an important factor (Denham, Bassett, & Wyatt, 2007; Eisenberg, Cumberland, & Spinrad, 1998). Parental socialization occurs both through global aspects of general parenting style (warmth, control) and through specific parenting practices (modeling, teaching, reactions to the child), and it is influenced by cultural factors (e.g., norms, values, gender stereotypes), parents’ characteristics (e.g., personality, emotional expressivity), children’s characteristics (e.g., temperament, age, gender), and the context in which parent-child interaction occurs (e.g., history of interactions) (Denham et al., 2007; Eisenberg et al., 1998). Parenting styles and practices involving negative appraisals of the child are likely to play a key role in the development of proneness to shame. Attributional perspectives suggest that socialization experiences and child characteristics that affect susceptibility to global self-attributions are likely to be key factors in the development of individual differences in shame (H. B. Lewis, 1971; M. Lewis, 1992; Weiner, 1986). Parenting is shame inducing when it precipitates global negative self-attributions either directly through global negative attributional statements to the child (e.g., criticism or scolding that blames essential characteristics of the child) or indirectly through actions or emotional expressions that signal disappointment, anger, disapproval, disgust, or contempt and generate such attributions by the child himself or herself (Campos, Thein, & Owen, 2004; M. Lewis, 1992). Love withdrawal, expressions of disgust or contempt, and humiliation are highly likely to activate global negative self-attributions and thus are considered prototypical forms of shaming (M. Lewis, 1992). Physical punishment and other forms of punitiveness also may be shame inducing if they are interpreted as reflecting hostility (Gershoff, 2002; Khaleque, & Rohner, 2002). Thus, by promoting global negative self- attributions, parental shaming is likely to contribute to the development of individual differences in proneness to shame. Research is sparse on the relations between shame responding in early childhood and parenting styles and practices. The influence of parents has not been examined, and the extent to which it may be conditional on child characteristics is unknown. In observational studies, more negative and less positive parental feedback about task performance was found to be associated with increased shame responses to task failure at age 3 (Alessandri & Lewis, 1993, 1996), and negative maternal feedback during a teaching task at 24 months predicted more shame at 36 months (Kelley et al., 2000). Global negative feedback did not predict shame responding in these studies, possibly because of the low rate at which it was observed. In a longitudinal study of young girls, a relatively authoritarian parenting style Shame in Early and Middle Childhood 503 (punitive, critical, unsupportive) by both mother and father when a child was age 3 predicted higher shame responding at age 5 (Mills, 2003). In a longitudinal study of young boys, on the other hand, boys whose mothers had been more intrusive (imposing their own goals) when the boys were toddlers were less prone to exhibit shame at age 3 (Belsky, Domitrovich, & Crnic, 1997). Given limited information regarding the socialization of shame responding in young children and how it may interact with other factors to promote shame, a major objective of the present study was to examine the prediction of shame responding in early childhood from parental shaming (verbal hostility, physical punishment, punitiveness). Mothers’ and fathers’ behavior both can contribute to the association between parenting patterns and shame responding. A growing body of research suggests that substantial parenting similarities and differences exist between mothers and fathers and that the interplay between the parenting styles and practices of both parents together may influence children’s development significantly (e.g., Lindsey & Mize, 2001; Nelson, Hart, Yang, Olsen, & Jin, 2006; Winsler, Madigan, & Aquilino, 2005). The finding that girls’ shame responding was predicted by the combined effect of both parents (Mills, 2003) suggests that the behavior of both mothers and fathers may interact to promote the development of shame responding. Thus, in the present study, it was of interest to determine whether high shaming by both mothers and fathers was associated independently or interactively with child shame responding. The influence of parental socialization also is likely to depend on child characteristics that affect the way children respond to parental behavior. There has been little attention to the extent to which links between parenting practices and shame responding are moderated by child characteristics that affect children’s emotional responses to parental behavior. Temperament is likely to be important (Lagattuta & Thompson, 2007; M. Lewis, 2007). A great deal of attention has been paid to the role in self-conscious emotions of fearful inhibition or “behavioral inhibition,” including low threshold of arousal, subjective feeling of anxiety or fear, and inhibited responses to novel, challenging, sudden, or intense events (Derryberry & Rothbart, 1997; Kagan, Reznick, & Snidman, 1987). Children with this temperament trait are more likely than other children to experience internal cues of discomfort or distress in response to parental discipline, and their bodily sensations are likely to facilitate self-focused attention and thus a tendency to make global negative self-attributions about their behavior (Dienstbier, Hellman, Lehnhoff, Hillman, & Volkenaar, 1975; M. Lewis, 1992). Thus, Kochanska and colleagues have demonstrated that reasoning and induction predict the development of conscience more strongly 504 Merrill-Palmer Quarterly among high-inhibited than low-inhibited children (Kochanska, 1991; Kochanska, DeVet, Goldman, Murray, & Putnam, 1994). By the same token, it has been suggested that high behavioral inhibition may magnify the effect of parental shame induction by facilitating or increasing self-focused attention (e.g., Lagattuta & Thompson, 2007; M. Lewis, 2007). To date, however, no research has investigated whether behavioral inhibition and parental shaming interact to influence the development of shame responding in early childhood. The present study examined whether parental shaming is more strongly related to shame for highly inhibited children than for other children. Another child characteristic that may affect the relation between parental shaming and children’s shame responding is the gender of the child. Girls may be more sensitive to shame induction than boys. Some evidence indicates that girls are more prone to shame than boys from an early age (Barrett et al., 1993; M. Lewis et al., 1992). There is also evidence for differential socialization of submissive and assertive emotions in girls and boys (Chaplin, Cole, & Zahn-Waxler, 2005; Crick & Zahn-Waxler, 2003), suggesting that girls may be more encouraged to experience or express shame than boys. There is limited evidence for gender differences in shame-promotive parenting, however. One study of parental evaluative feedback in relation to shame in 3-year-olds (Alessandri & Lewis, 1993) found that parents gave girls more negative and less positive feedback about task performance. To the extent that girls are exposed to more shaming than boys, and thus are more sensitive to shame, parental shaming may be more strongly related to shame for girls than for boys. Thus, we examined whether relations between shame responding and parental shaming were different for girls and boys. Another objective of the study was to determine whether shame responding assessed in early childhood from emotion-expressive responses was related to shame proneness assessed in middle childhood. In early childhood, children’s proneness to shame was assessed from observations of their reactions to failing performance tasks designed to manipulate attributions about responsibility for the outcome (Alessandri & Lewis, 1996; Belsky et al., 1997; Kelley et al., 2000; M. Lewis et al., 1992; Mills, 2003). Although the measure involved, by necessity for ethical reasons, a mild shame induction, more pronounced shame responding to a mild eliciting situation seems potentially nonnormative and a reflection of proneness to shame responding. Proneness to shame responding, if reinforced by shaming experiences, would strengthen a tendency to respond with shame and facilitate the development of a disposition to this emotion. Thus, it was of interest to establish whether shame responding to the failure paradigm Shame in Early and Middle Childhood 505 would predict shame proneness in middle childhood. We expected that the likelihood of such a relation would be greater for those children exposed to parental shaming. In summary, this study addressed several questions about the sources of early individual differences in shame responding that have received little or no attention previously: (a) whether shame responding is related to parental shaming, and whether mothers and fathers have independent and/or interactive effects; (b) whether child temperamental inhibition moderates associations between shame responding and parental shaming; and (c) whether child gender moderates associations between shame responding and parental shaming. A second objective was to determine whether shame responding to the failure paradigm in early childhood predicts proneness to shame in middle childhood among children exposed to parental shaming. Parenting was assessed both by parents’ self reports and by their partners’ reports of their parenting practices. Self-reports and spouse reports tend to show modest agreement (Russell, Hart, Robinson, & Olsen, 2003; Winsler et al., 2005). In some research, spouse ratings of parenting behavior have been used in addition to or in place of self-ratings in an effort to avoid social desirability effects (Nelson et al., 2006; Nix et al., 1999; Porter et al., 2005; Russell et al., 2003; Yang et al., 2004). In some of these studies, spouse reports appeared to yield more meaningful findings (Yang et al., 2004), whereas in others more relations between parenting and child outcome were found when self-reports rather than spouse reports were used (Russell et al., 2003). To increase the reliability of measurement, both selfreports and spouse reports were obtained in the present study. Method Participants Participants were 225 children (128 boys, 97 girls) and their parents (218 mothers, 184 fathers) involved in a longitudinal study that began when the children were between 3.7 and 4.5 years of age (M = 4.14 years, SD = 0.24). Of the 225 children who participated at the first time point (Time 1), 199 (88%) participated at the second time point (Time 2) when they were between 5.3 and 7.3 years of age (M = 5.89 years, SD = .26), and 162 (72%) participated at the third time point (Time 3) when they were between 7.6 and 9.4 years of age (M = 8.15 years, SD = .26). Families with young children were recruited through a letter of invitation under a cover letter sent by a government agency responsible for 506 Merrill-Palmer Quarterly administering health care. The agency generated a mailing list of 3,500 families drawn randomly from all 6,358 families residing in an urban area (pop. 600,000) as of June 1, 2002, with a child born between June 1, 1999, and May 31, 2000, and therefore 3 or 4 years of age at the time of recruitment. To protect their privacy, recipients of the letter were unknown to the researchers unless they self-identified as interested in participating in the study. Recruitment began as soon as the mailing list was generated, and was conducted in three stages over an 18-month period until the target sample size was reached. At each stage, letters of invitation were sent to approximately one third of families on the mailing list. Details were provided about the study, and families were invited to return an enclosed stamped return postcard if they had a healthy child and wished to participate. Of 257 families volunteering to participate, 27 served in the pilot phase of the study, 4 withdrew during the first wave, and 1 was excluded because of a lack of knowledge of English, leaving 225 families. The recruitment procedure did not allow the computation of a response rate because it could not be determined how many families were eligible and how many eligible families had moved and failed to receive the invitation by the time of the mailing. Children were predominantly firstborns (36%) or secondborns (36%), with 5% being singletons. Parents predominantly were married or cohabiting (90% of mothers, 97% of fathers), were in their 30s or 40s (83% of mothers, 81% of fathers), had a postsecondary trades or community college certificate (49% of mothers, 42% of fathers) or a university degree (35% of mothers, 31% of fathers), and were employed (93% of fathers, 41% of mothers; 59% of mothers were employed part time). An assessment of socioeconomic status from education and occupation (Hollingshead, 1975) indicated that most families (70%) were middle class (skilled worker or medium business/professional), with 16% being lower (semiskilled or unskilled) and 14% higher (major business/professional). Families that continued to participate at Time 2 (n = 200) did not differ significantly from those that withdrew (n = 29) on Time 1 measures of parental behavior, child temperamental inhibition, shame responding, or most demographics, but participants who continued were higher in socioeconomic status, t(219) = 3.04, p < .01, had a higher family income, t(215) = 3.71, p < .001, were more likely to be married or cohabiting, χ2(1, N = 223) = 9.52, p <.01, and were more likely to have a son than a daughter in the study, χ2(1, N = 225) = 4.07, p < .05. Mothers in families that continued to participate at Time 3 (n = 162) had a higher level of education than those in families that withdrew (n = 37), χ2(1, N = 196) = 4.31, p < .04. Shame in Early and Middle Childhood 507 Overview Parents provided informed consent for their own and their child’s participation in the procedures. At each time point, children came to the laboratory for a visit, which took place in a playroom equipped with a remote-controlled camera operated from behind a one-way mirror. During a familiarization period following arrival, the experimenter engaged the child in casual conversation about school and siblings while making things with Play Doh and doing a jigsaw puzzle together. Then, to measure shame responding at Time 1 and Time 2, observations were made of children’s emotion-expressive behavior in response to failure on tasks used in previous work (Geppert, 1986; M. Lewis & Ramsay, 2002; Mills, 2003; Stipek, Recchia, & McClintic, 1992). During the assessment, the parent completed questionnaires at a table some distance away and behind the child (Time 1) or in another room (Time 2). At Time 3, children were administered a series of measures in a structured interview format, one of which assessed shame proneness. Prior to lab visits, parents had been mailed a package of questionnaires to complete that included measures of their parenting practices and their child’s temperamental characteristics. Coparents were asked to complete the questionnaires independently of one another. Time 1 and Time 2 Responses to Failure Children’s shame responding in early childhood was assessed by using failure situations validated in past work on children’s shame-expressive behavior (Alessandri & Lewis, 1996; Belsky et al., 1997; Kelley et al., 2000; M. Lewis et al., 1992; Mills, 2003). According to functionalist theories, an emotion state can be inferred from patterns of expressive behavior that reflect the action tendency associated with it occurring in a context appropriate to it (e.g., M. Lewis & Michalson, 1983). In the present study, shame responding was assessed from emotion-expressive behavior reflecting the action tendency of shame (hiding or avoidance; e.g., averting gaze) occurring in response to failing performance tasks that might induce shame by precipitating a global negative self-attribution about the outcome. Other potential emotional reactions to failure were also assessed. Time 1. Children were seated at a child’s table facing a small digital remote-controlled camera placed in an unobtrusive location about 10 feet away. The (female) experimenter was seated to the children’s left on the adjacent side of the table. To assess individual differences in shame responding, performance tasks were administered in a fixed order, as follows: (a) 508 Merrill-Palmer Quarterly two versions of a matching task, the first portrayed as easy (“Most children your age could finish in time”) and the second as difficult (“Most children your age could not finish in time”); (b) two versions of a competitive nesting task, the first portrayed as easy and the second as difficult; and (c) two easy tasks that were demonstrated to be easy but then were surreptitiously modified to prevent success, the first a simple five-piece puzzle with one piece replaced by another that had been slightly altered so it would not fit and the second a set of small cups for nesting that was altered by replacing one cup with another that was the same size as one of the other cups. The outcome of each performance was controlled to ensure success and then failure on each version of the matching and competitive nesting tasks, and failure on each of the impossible tasks, resulting in a total of six failure experiences. Easy tasks were included because shame is more likely to occur in response to failing an easy than a difficult task (M. Lewis et al., 1992), because of the greater likelihood of a negative self-attribution about failure when the task is easy. Difficult tasks were included because they might be more shame inducing for children highly prone to shame. Success experiences were included to avoid undue distress and to keep children engaged. After drawing attention to the outcome (e.g., “Time is up, you did not finish before the bell”), the experimenter remained silent and still for an interval of time (e.g., 30 sec). Emotional displays observed during this time were coded to measure emotional responses to failure (see the subsection Coding of emotion-expressive behaviors). At the conclusion of the impossible tasks, the experimenter looked through her materials and said, “Oh, no, I made a mistake. I gave you the wrong pieces for the puzzle and cups. No one can do the puzzle with these pieces, and the cups won’t fit because two of them are the same size.” The experimenter then asked whether the child would like to work on a puzzle or set of cups that would not be impossible to do and praised the child’s performance. Finally, the experimenter complimented the child’s performance on the matching and nesting tasks and, to dispel any lingering distress associated with the failure experiences, noted problems with the timer that could explain why the child had sometimes not finished in time. Time 2. As at Time 1, children were seated at a child’s table while facing a digital remote-controlled camera with the (female) experimenter seated to the children’s left on the adjacent side of the table. They were given three timed puzzles to complete. Each was an eight- or nine-piece puzzle of well-known characters doing simple activities (e.g., Barney playing football) that was designed for 3- to 4-year olds. The experimenter set a time limit for each puzzle, and timed the child with a stopwatch. After showing the intact puzzle to the child, the experimenter took the pieces out and said, “You should be able to do this one in 2 minutes, no problem. Shame in Early and Middle Childhood 509 Ready, set, go!” When the child was within two pieces of completing the puzzle, the experimenter pressed the stop button and announced “Time’s up.” Two puzzles were administered this way. To alleviate distress, a third puzzle was done with the time limit set so that the child would succeed. Then, the experimenter checked her notes, discovered that she had been setting time limits way too short, apologized, and said, “No one could do these puzzles that fast. I’ll bet you could’ve done all of them in time if I’d given you the right amount of time.” Coding of emotion-expressive behaviors. Children’s emotion-expressive behaviors occurring during the interval following task failure were coded moment by moment for facial or bodily cues relevant to shame. At Time 1, two other potential emotional reactions to failure—sadness and anger— also were coded to assess the specificity of the emotional responses to failure. Based on previous work (Barrett et al., 1993; Geppert, 1986; Izard, 1979; Keltner, 1995; M. Lewis et al., 1992; Tracy & Matsumoto, 2008; Tracy, Robins, & Schriber, 2009), shame was assessed from gaze averted or lowered (not attributable to a meaningful event), head tilted forward and down, body collapsed or slumped (shrinking or compression of trunk), and hiding of one’s face (moving one’s hands to cover the face or part of the face, or moving the face or head to cover the face). Sadness and anger were defined on the basis of prototypical facial cues, any one or more of which could signify the emotion (Cole, Barrett, & Zahn-Waxler, 1992; Cole, Zahn-Waxler, & Smith, 1994; Izard, 1979). For sadness, the cues were lip corners turned down, eyelids drooped, inner brows raised and/or lowered (/ \), and head dropped and tilted to the side. For anger, the cues were eyes narrowed, brow furrowed, mouth set with lips pressed/tightened, and teeth clenched (lower teeth visible). The initial occurrence of each cue was rated on a 3-point intensity scale: 0 = not clearly visible or absent; 1 = mild to moderate intensity; 2 = extreme intensity. Coders were the first author and three advanced graduate students unaware of the goals of the study and without knowledge of the specific signs that would be used to score each emotion. Coders were trained to a high level of reliability on ratings, conducted reliability checks regularly, and resolved disagreements through discussion. Ratings of each cue were averaged across failure experiences to create a single overall rating. Reliability was calculated based on 20% of the actual cases. Intraclass correlations yielded interrater agreement estimates for shame cues ranging from .86 to .96 at Time 1 and from .84 to .99 at Time 2. Time 1 sadness and anger cues occurred infrequently, and intraclass correlations could not be calculated. Percentage agreement for these cues ranged from .87 to .98 for sadness and from .83 to .98 for anger. A measure of shame at 510 Merrill-Palmer Quarterly each time point and measures of sadness and anger at Time 1 were computed by averaging the ratings for the component cues (possible range, 0–2) and computing the mean across failure experiences to create a single overall score for each emotion. Time 1 Parenting Parents were asked to independently complete the Parenting Styles and Dimensions Questionnaire (PSDQ; Robinson, Mandleco, Olsen, & Hart, 1995, 2001) by reading each statement and rating on a 5-point scale the frequency with which their spouse, and then they themselves, exhibited the behavior described in the statement: 1 = never; 2 = once in a while; 3 = about half of the time; 4 = very often; 5 = always. The PSDQ was developed with parents of preschool-age and school-age children. It yields three global factors (authoritative, authoritarian, permissive) and internally consistent subscales for both mothers and fathers. It has good factorial validity and predictive validity as indicated by relations with child social behavior outcomes (Russell et al., 1998). Parental shaming was measured by the 12 items of the authoritarian factor assessing verbal hostility, physical punishment, and punitiveness (alphas were .78 for mother self-reports, .75 for father self-reports, .82 for father reports about mother, and .82 for mother reports about father). These reliabilities were comparable to those found in previous research (Robinson et al., 2001; Winsler et al., 2005). Agreement between self-reports and spouse reports was assessed by computing intraclass correlations between parents’ self-reports and their partners’ reports about them. Correlations were .52 and .68 for self-reports and spouse reports of mother shaming and father shaming, respectively. Correlations between parents’ self-reports and reports on their partner were .75 and .78 for mother shaming and father shaming, respectively. It appeared that parent ratings of self and partner were highly concordant, suggesting some confounding between self-perceptions and partner perceptions. Given the relatively low agreement between self-reports and spouse reports for mother shaming, analyses were done with self-reports and spouse reports separately. Time 1 Child Temperament Child temperamental inhibition was assessed by using the Children’s Behavior Questionnaire (CBQ; Rothbart, Ahadi, & Hershey, 1994; Rothbart, Ahadi, Hershey, & Fisher, 2001), a parent-report measure comprised of Likert-scaled items describing children’s reactions to situations. Parents Shame in Early and Middle Childhood 511 indicate how well each statement describes their child by using a 7-point scale: 1 = extremely untrue of your child; 2 = quite untrue; 3 = slightly untrue; 4 = neither true nor false; 5 = slightly true; 6 = quite true; 7 = extremely true. The CBQ yields internally consistent scales that are stable over time and have convergent validity as indicated by parental agreement (Rothbart et al., 2001). Temperamental inhibition was assessed by averaging the items belonging to the fear, shyness, and discomfort subscales (Kochanska et al., 1994); alphas were .89 for mothers and .86 for fathers. Mother and father reports were correlated, r(127) = .64, p < .001, and were averaged to create a single overall measure. Time 3 Proneness to Shame Children’s shame proneness was assessed by using the Test of Self-Conscious Affect for Children (TOSCA-C; Tangney, Wagner, Burggraf, Gramzow, & Fletcher, 1990), a scenario-based measure comprised of potentially shameinducing and/or guilt-inducing everyday situations (e.g., making a mistake, damaging someone’s property, failing at something), each followed by situation-specific phenomenological descriptions of affective, cognitive, or behavioral responses capturing shame (negative self-feeling or judgment, desire to hide), guilt (remorse, repair), externalization (blaming the other person or the situation), detachment (unconcern), pride in one’s self, and pride in one’s behavior. Respondents imagine themselves in each situation and rate the likelihood of each response on a 5-point scale: 1 = not likely; 5 = very likely; for example, “Your aunt is giving a big party. You are carrying drinks to people, and you spill one all over the floor: (a) you would think ‘I should have been more careful;’ (b) you would think ‘my aunt wouldn’t mind that much’; (c) you would run upstairs to be away from everybody; (d) you would think the tray was too heavy.” The TOSCA-C has good internal consistency (Tangney & Dearing, 2002; Tangney et al., 1990). Shame and guilt scores were computed from children’s responses to 10 negative scenarios. Responses were averaged across scenarios; alphas were .78 and .74 for shame and guilt, respectively. Because shame and guilt are positively correlated but related in opposite directions to other variables (Tangney, 1995), they are believed to operate as cooperative suppressors (Paulhus, Robins, Trzesniewski, & Tracy, 2004; Tangney, 1995). As has been repeatedly demonstrated (Tangney, 1995; Tangney & Dearing, 2002), their power to predict other variables improves when their common variance is removed. To analyze shame unconfounded by guilt, shame proneness was measured by computing a residual shame score in which the variance attributable to guilt was removed. 512 Merrill-Palmer Quarterly Results Preliminary Analyses An examination of missing values revealed that the proportion of missing data was less than 5% for parent reports of child temperament and mother self-reports of shaming. Higher proportions of missing data were found for children’s responses to failure (7.6% missing at Time 1, most often because of withdrawal from the task; 13.3% missing at Time 2), mother reports about fathers (8.9% because of a small proportion of single-parent mother-headed families in the sample), and father reports about themselves and their partner (18.7% missing because of fathers’ lower participation rate). At Time 3, a substantial proportion of data (32.4%) was missing for the assessment of children’s proneness to shame. Little’s test of data missing completely at random indicated that the pattern of missing values deviated significantly from complete randomness by Time 3, ps =.07, .16, and .01 for Times 1, 2, and 3, respectively. Cases with and without missing data did not differ significantly on any of the study variables at Times 1 or 2. Cases with missing data for the assessment of proneness to shame at Time 3 showed less sadness in response to failure at Time 1, t(147.4) = 3.20, p < .01, and lower self-reported father shaming at Time 1, t(111.6) = 2.10, p = .04. These findings indicate that missingness was random with respect to all but one of the key variables of interest. To reduce bias and loss of power because of missing values, the analyses were performed with missing values imputed. Single imputation was performed by using maximum likelihood estimation. Prior to analysis, the variables were examined. Most distributions were normal. A small number of outlying values were recoded to the highest or lowest value that was not an outlier (Tabachnick & Fidell, 2001). Mother and father shaming scores were positively skewed and were subjected to log transformations. Table 1 lists descriptive statistics for the variables in their original form. There was substantial variation in the intensity of shame responding, indicating that the shame assessment measured individual differences in shame responding. Parental shaming was quite low, occurring on average less than “once in a while” (2 on a 5-point scale). Emotional reactions to failure. Shame responding at Time 1 was not strongly related to sadness or anger for girls, rs(95) = .12 and .02, respectively, and was significantly related to sadness but not to anger for boys, r(126) = .18, p < .05, and r(126) = .03, respectively. Given that failure was the context in which the emotions were assessed, the low magnitude of the correlations indicates that shame was the more prominent reaction. Shame in Early and Middle Childhood 513 Table 1. Descriptive Statistics Girls Boys Mean SD Mean SD Time 2 shame 0.98 0.23 0.89 0.27 Time 1 shame 0.75 0.20 0.73 0.18 Time 1 sadness 0.11 0.09 0.11 0.09 Time 1 anger 0.14 0.09 0.15 0.10 Mother self-report 1.74 0.41 1.82 0.38 Mother spouse report 1.76 0.45 1.75 0.42 Father self-report 1.86 0.41 1.90 0.40 Father spouse report 1.78 0.56 1.86 0.44 3.72 0.70 3.80 0.64 Shame 2.60 0.60 2.24 0.70 Guilt 3.96 0.50 3.63 0.68 Shame proneness (residual score) 0.10 0.55 –0.09 0.60 Responses to Failure Time 1 Parental Shaming Time 1 Temperament Temperamental inhibition Time 3 Shame and Guilt Gender differences. At Time 1, girls and boys did not differ significantly in age, t(223) = 0.52 (d = .07), temperamental inhibition, t(223) = –0.83 (d = .11), or experiences of parental shaming, ts(223) = 1.76 and 1.82, ps < .07 (both ds = .24) for mother reports of self and partner, respectively, and ts(223) = .73 (d = .04) and .03 (d = .00) for father reports of self and partner, respectively. Girls and boys did not differ significantly in emotional responding to failure, ts(223) = 0.96, –.08, and –1.37 (ds = .13, .01, and .19) for shame, sadness, and anger, respectively. Girls showed more shame responding to failure than did boys at Time 2, t(223) = 2.40, p < .02 (d = .33), and scored higher than did boys on shame proneness at Time 3, t(223) = 2.41, p < .02 (d = .33). Parent differences. Differences between mothers’ and fathers’ perceptions of their own and each other’s parenting were examined. Analyses of parent self-perceptions indicated that fathers rated their use of shaming higher than did mothers, t(96) = –2.97, p < .01 (d = .35) for parents of girls and t(127) = –2.05, p < .04 (d = .22) for parents of boys. Comparisons between parents’ ratings of themselves and their partner (self-reports and 514 Merrill-Palmer Quarterly reports about partner) revealed the same differences, with fathers rating themselves higher on shaming than they did their partner, t(96) = 4.22, p < .001 (d = .36) for fathers of girls and t(127) = 6.88, p < .001 (d = .48) for fathers of boys. These findings are in line with previous research (Winsler et al., 2005) and indicate that mothers and fathers perceive differences between themselves and their partners, viewing mothers as using less shaming than fathers. Prediction of Shame Responding to Failure in Early Childhood The hypothesis that shame responding would be associated with parental shaming, and that child temperamental inhibition would moderate these associations, was evaluated by using sequential regression for each gender to determine whether the addition of temperament and parenting × temperament interactions would improve the prediction of shame responding beyond that afforded by parenting practices. In each analysis, the overall shame score was predicted from parental shaming (maternal, paternal) entered first (Step 1), followed by the interaction of maternal and paternal shaming (Step 2), child temperamental inhibition (Step 3), the two-way interactions between parental shaming and temperamental inhibition (Step 4), and the three-way interaction among mother shaming, father shaming, and temperamental inhibition (Step 5). To provide for more meaningful interpretation of the solutions, the predictors were rescaled by putting them in the form of mean-deviation scores (centering) (Aiken & West, 1991; Whisman & McClelland, 2005). Where interactions were found, the simple slopes were tested by following recommended procedures in which the regression is performed at high (+1 SD) and low (–1 SD) values of the putative moderator variable (Aiken & West, 1991). As Step 1, zero-order correlations among the variables were computed (see Table 2). Child age at Time 1 was included to determine whether age was associated with any of the Time 1 variables and could be confounded with their associations. There were few significant correlations among the key study variables. These findings supported the objective of the present study to examine shame responding in relation to the interactive effects of child characteristics and parenting. Shame responding showed modest stability from Time 1 to Time 2 for girls, r(95) = .19, p = .057. Child age was correlated with some variables. For girls, age was significantly correlated with higher Time 1 shame responding, r(95) = .33, p < .001, but not with other measures, largest rs(95) = .14, for spouse reported father shaming. For boys, age was associated with higher self-reported father shaming, rs(126) = .20, p < .03, spouse-reported father shaming, rs(126) = .18, p < .05, and .33*** 6. T1 Father shaming: self .20* .05 9. T3 shame proneness .39*** .23* –.14 .59*** .04 –.05 .02 –.02 .23** .14 –.13 .06 7 .54*** .29** .65*** –.05 –.05 .00 8 –.07 .01 –.01 –.04 –.10 –.24* –.03 .13 .09 .53*** .69*** .64*** .32*** .11 –.11 –.03 6 *p < .05. **p < .01. ***p < .001. Note. T1 = Time 1; T2 = Time 2; T3 = Time 3. Correlations are above the diagonal for girls and below the diagonal for boys. –.02 .08 8. T1 Temperamental inhibition .10 .03 –.23** .09 7. T1 Father shaming: spouse .18* .02 .27** 5 –.05 .33*** .25** 5. T1 Mother shaming: spouse .17 .01 .00 –.02 –.03 –.02 4 –.03 4. T1 Mother shaming: self .06 .05 3. T1 shame responding .15 .07 .19 3 .33*** 2 2. T2 shame responding .07 1 1. T1 child age .00 Table 2. Intercorrelations –.10 –.04 –.22* –.16 –.23* –.21* .11 –.05 9 Shame in Early and Middle Childhood 515 516 Merrill-Palmer Quarterly Table 3. Concurrent Prediction of Time 1 Shame Responding Girls B SE B Boys β B SE B β –.12 Mother shaming –.04 .21 –.02 –.25 .19 Father shaming .35 .23 .16 .68 .18 Mother × Father .34*** –3.52 2.61 –.14 –1.40 1.94 –.06 Temperamental inhibition –.02 .03 –.07 .02 .02 .05 Mother × Inhibition –.17 .32 –.06 .54 .31 .15 Father × Inhibition .01 .31 .00 –.02 .28 –.01 Mother × Father × Inhibition –3.93 3.64 –.11 –1.88 3.26 –.05 R² = .17 (adj. = .09) R² = .17 (adj. = .11) F(8, 88) = 2.21, p < .03 F(8, 119) = 2.94, p < .01 Note. Analyses are adjusted (adj.) for child age at Time 1 (β = .33, p < .001 for girls; β = .09 for boys). ***p < .001. spouse-reported mother shaming, rs(126) = .17, p < .05, but not with Time 1 shame responding, rs(126) = .15, p < .10, or self-reported mother shaming, rs(126) = .06. Because of its associations with some variables, child age was included as a covariate in the analyses. Concurrent prediction of shame responding at Time 1. Two regressions were performed to assess the concurrent prediction of shame responding at Time 1 from self-reported parental shaming and child temperamental inhibition. The results after entry of all predictors are listed in Table 3. For girls, the sole significant associate of shame responding was age, β = .33, p < .001 (sr2 = .11, indicating 11% unique variance explained). For boys, the sole significant associate was father shaming, β = .34, p < .001 (sr2 = .10). The correlation found between boys’ age and father shaming indicates that age was indirectly associated with boys’ shame responding. Separate analyses performed by using spouse reports of parenting yielded several independent or interactive effects of parental shaming. Girls’ shame responding was significantly associated with father reports of higher mother shaming, β = .36, p < .001 (sr 2 = .11), mother reports of lower father shaming, β = –.19, p < .06 (sr 2 = .03), and an interaction between father reports of mother shaming and mother reports of father shaming, β = –.22, p < .03 (sr 2 = .04). Simple slope tests on the interaction indicated that higher mother shaming as reported by fathers was significantly associated with higher shame responding for girls whose fathers Shame in Early and Middle Childhood 517 Table 4. Longitudinal Prediction of Time 2 Shame Responding From Time 1 Predictors Girls B Shame responding Mother shaming Father shaming Boys SE B β .17 .11 .17 .10 .21 .05 –.11 –.27** B SE B β .10 .15 .07 .29 .30 .09 –.16 .31 –.05 –.29 3.13 –.01 –.25 .24 –7.29 2.71 Temperamental inhibition –.06 .03 –.23* .00 .04 .01 Mother × Inhibition –.33 .32 –.11 .94 .50 .17 Father × Inhibition .21 .32 .07 .16 .45 .03 Mother × Father × Inhibition .04 3.75 .00 9.56 5.24 .17 Mother × Father R² = .19 (adj. = .10) R² = .08 (adj. = .01) F(9, 87) = 2.19, p < .03 F(9, 118) = 1.08, ns Note. Analyses are adjusted (adj.) for child age at Time 1 (β = –.07 for girls; β = .05 for boys). *p < .05. **p < .01. were low in shaming according to mother reports, β = .53, p < .001, but not for girls whose fathers were high in mother-reported shaming, β = .17, ns. Boys’ shame responding was significantly associated with father reports of higher mother shaming, β = .24, p < .01 (sr 2 = .05). Longitudinal prediction of shame responding at Time 2. Two regressions were performed to assess the prediction of Time 2 shame responding longitudinally from Time 1 shame responding, self-reported parental shaming, and temperamental inhibition. The results after entry of all predictors are listed in Table 4. Girls’ shame responding was predicted by lower temperamental inhibition, β = –.23, p < .03 (sr 2 = .05), and by a twoway interaction between mother and father shaming, β = –.27, p < .01 (sr 2 = .07). Slope tests on the interaction indicated that higher mother shaming predicted increased shame responding by school age if father shaming was low, β = .33, p < .04, but not if it was high, β = –.24, ns. Thus, for girls, lower temperamental inhibition, or higher mother shaming occurring in conjunction with low father shaming, predicted increased shame responding at school age. Boys’ shame responding was predicted by a marginally significant twoway interaction between mother shaming and temperamental inhibition, β = .17, p = .061 (sr 2 = .03). Planned slope tests on the interaction revealed 518 Merrill-Palmer Quarterly that mother shaming predicted increased shame responding at school age for high-inhibited boys, β = .25, p < .07, but not for low-inhibited boys, β = –.12. The analyses were also performed by using spouse reports of parenting rather than self-reports. No significant effects were found by using spouse reports for either girls or boys. Prediction of Shame Proneness in Middle Childhood Two regressions were performed to assess the prediction of Time 3 shame proneness from Time 1 shame responding, self-reported parental shaming, and their two- and three-way interactions. The results of the analyses after entry of all predictors are shown in Table 5. Girls’ shame proneness was predicted by a two-way interaction between shame responding and father shaming, β = .22, p < .06 (sr 2 = .03), and a three-way interaction among shame responding, mother shaming, and father shaming, β = –.29, p < .01 (sr 2 = .06). Slope tests on the three-way interaction are shown in Figure 1. Higher shame responding at preschool age predicted lower proneness to shame in middle childhood for girls with low-shaming mothers and fathers, β = –.50, p < .01, and to a lesser extent, higher proneness to shame for those with highshaming fathers but low-shaming mothers, β = .57, p = .071. Slopes were far from significant for girls with high-shaming mothers and fathers, β = –.24, and girls with high-shaming mothers but low-shaming fathers, β = –.14. Boys’ proneness to shame in middle childhood was predicted by lower father shaming, β = –.19, p < .06 (sr 2 = .03), and by an interaction between mother and father shaming, β = .22, p < .05 (sr 2 = .04). Simple slope tests on the interaction showed that mother shaming predicted higher subsequent proneness to shame for boys whose fathers were high shaming, β = .32, p < .01, but not for those whose fathers were low shaming, β = –.11, ns (Figure 2). Separate analyses performed by using spouse reports of parental shaming yielded a significant interaction effect for girls between shame responding and mother-reported father shaming at preschool age, β = .28, p < .03 (sr 2 = .05), indicating that greater shame responding at preschool age predicted lower proneness to shame in middle childhood when fathers were low in mother-reported shaming, β = –.36, p < .01, but not when they were high in mother-reported shaming, β = .06, ns. There were no significant effects for boys involving spouse-reported shaming. Discussion The goals of this study were to examine individual differences in proneness to shame in early childhood and assess their predictive relations with F(8, 88) = 2.57, p < .01 β –.29** –.06 .22 –.11 –.17 –.16 –.08 *p < .05. **p < .01. .45 .44 B 34.18 5.75 2.68 3.08 .50 .53 .26 SE B Boys F(8, 119) = 2.59, p < .01 R² = .15 (adj. = .09) 26.91 13.03 –1.22 2.91 –.97 Note. Analyses are adjusted (adj.) for child age at Time 1 (β = –.03 for girls; β = –.19, p < .03 for boys). R² = .19 (adj. = .12) 6.65 35.91 –3.63 –91.31 Mother × Father Shame responding × Mother × Father 3.72 3.06 7.18 .58 .53 –3.07 –.97 Father shaming Shame responding × Father –.83 Mother shaming .30 SE B Girls Shame responding × Mother B –.19 Shame responding Table 5. Prediction of Time 3 Shame Proneness From Time 1 Predictors β .08 .22* –.04 .09 –.19 .08 .17 Shame in Early and Middle Childhood 519 520 Merrill-Palmer Quarterly 1 High MS, High FS Time 3 Shame Proneness, 0.9 High MS, Low FS 0.8 Low MS, High FS Low MS, Low FS 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Low High Time 1 Shame Responding Figure 1. Girls’ Time 3 shame proneness as predicted by a three-way interaction among Time 1 shame responding to failure, mother shaming, and father shaming. MS = mother shaming; FS = father shaming. proneness to shame in middle childhood. Girls showed more shame than boys by school age. For girls, being older was associated with more shame responding at preschool age. Girls who were low in inhibition or experienced higher shaming specifically by mother showed increased shame responding between preschool age and school age. For boys, father shaming was associated with more shame responding at preschool age. Highinhibited boys who experienced higher mother shaming tended to show increased shame responding between preschool age and school age. Girls who showed more shame responding at preschool age were less shame prone in middle childhood if both parents were low shaming, and these girls were inclined to be more shame prone if they experienced high father shaming. Boys who experienced more mother and father shaming at preschool age were more prone to shame in middle childhood. Shame in Early and Middle Childhood 521 Time 3 Shame Proneness 2.5 2 1.5 1 Low Father Shaming 0.5 High Father Shaming 0 Low Mother Shaming High Figure 2. Boys’ Time 3 shame proneness as predicted by a two-way interaction between mother and father shaming. The findings confirm the existence of gender differences in the development of shame responding. Girls’ shame responding at preschool age was strongly age related, suggesting that this was a period of rapid development of shame. This is consistent with the notion that shame is valued in girls and is promoted early. Increases in shame responding as girls reached school age were predicted by low inhibition, perhaps because of a developmental lag in low-inhibited girls. Higher mother shaming was another predictor of increased shame responding in girls. This finding is consistent with attributional perspectives suggesting that certain child characteristics and parenting practices will facilitate internal attributions and promote global negative self-attributions about behavior (e.g., Dienstbier et al., 1975; Kochanska, 1991; Kochanska et al., 1994; Kochanska, Gross, Lin, & Nichols, 2002). Parental shaming is likely to promote global negative self-attributions by directly or indirectly conveying messages of hostility and blame, and girls may be particularly susceptible to messages that promote self-attributions, such that mother shaming by itself was sufficient to promote increased shame responding. Interestingly, the relation between higher mother shaming and 522 Merrill-Palmer Quarterly increased shame responding was found only in conjunction with low father shaming. High shaming by both mother and father was associated with girls’ lower shame responding. These findings seem to suggest that when girls are low shame responders both parents use more shaming. Boys’ shame responding at preschool age was associated solely with father shaming. Given that fathers used more shaming with older boys, it appears that fathers’ developmental expectations for their sons may have contributed to their use of shaming. In light of the finding that fathers used more shaming than mothers, it seems likely that fathers’ use of shaming plays an important role in promoting boys’ shame responding at preschool age. At school age, boys who had been high in behavioral inhibition and exposed to more mother shaming showed a greater increase in shame responding than other boys. As attributional perspectives suggest, parenting practices that promote global negative self-attributions are more likely to do so in the context of certain temperamental characteristics. When both temperament and parenting were conducive, boys’ shame responding was more likely to increase by school age, suggesting that high inhibition may have served to magnify the effect of mother shaming, possibly by increasing negative self-focused attention. That the effect fell short of significance in the present normative sample is consistent with the idea that boys may be relatively insensitive to shame induction. A second objective of the study was to assess predictive relations between early childhood shame responding and proneness to shame in middle childhood. Over time, the negative messages conveyed by parental shaming may engender a bias toward global negative self-attributions leading to a disposition to shame reactions. The present findings are consistent with such a developmental trajectory and also reveal gender differences. Girls who had shown more shame responding in early childhood were less shame prone in middle childhood if both parents had been low shaming and, to a lesser extent, more shame prone if they had experienced high father shaming. Thus, shame responding in early childhood appears to be a protective factor for girls when both mother and father shaming is low, and it may be a risk factor when father shaming is high. These findings suggest that shame responding is a valued and adaptive characteristic in girls and hence not by itself a risk factor for subsequent proneness to shame. For boys, as well, high levels of shaming predicted proneness to shame in middle childhood, but only when both parents were high shaming. This finding provides further support for the idea that boys may be relatively insensitive to shame and hence are likely to become prone to shame only when both parents are high shaming. Given research indicating that proneness to shame is associated with a wide range of psychological symptoms Shame in Early and Middle Childhood 523 by middle childhood (Ferguson, Stegge, Eyre, Vollmer, & Ashbaker, 2000; Ferguson et al., 1999) and potentially earlier (Luby et al., 2009), further longitudinal investigation is warranted to examine the developmental emergence of these links and possible differences between girls and boys. Self-reports and spouse reports of parenting did not yield the same results. Few effects were found by using spouse reports, and the effects were primarily for girls. Two of these effects were generally consistent with the results of analyses performed with self-reports: Mother shaming predicted greater shame responding in early childhood for girls who had low-shaming fathers, and early-childhood shame responding predicted lower proneness to shame for girls who had low-shaming fathers. Other effects could be interpreted as reflecting perceptions of self rather than of spouse. For example, boys’ higher shame responding to failure at Time 1 was associated with fathers’ ratings of mother shaming, mirroring the association found with fathers’ self-ratings of shaming. Given the correlations showing that self-ratings and ratings of partner agreed more closely than self-ratings and ratings by partner, it appears that parents may have assessed their partners in part on the basis of their own self-perceptions such that spouse reports were confounded with self-perceptions. To the extent this was so, the results of the analyses performed by using self-reports may be more meaningful than those performed by using spouse reports. Analyses using self-reports also yielded more coherent results. Several limitations of the study should be noted. Given that the results were based on parents’ self-reports of parenting, the findings may reflect parents’ perceptions of their parenting rather than their actual parenting. For example, the finding that mother shaming predicted girls’ shame responding could reflect a tendency for mothers of shame-prone girls to perceive their parenting practices as somewhat harsh. In future research, the use of other methods of assessing parenting practices would provide a more independent measure of parental shaming and help to substantiate the present findings. Another limitation of the study is some restriction in generalization of the findings, owing to the disproportionate loss of families with lower socioeconomic status following the first phase of the study and families with lower shaming fathers by the third phase. Notwithstanding these limits, the coherence of the results that emerged despite the low levels of parental shaming in the sample, the mild nature of the shame-induction procedures in early childhood, and the long time span of the study, lends credence to the findings. The study contributes new information regarding the sources of individual differences in shame responding in early childhood and its predictive relations with proneness to shame in middle childhood. First, the data reveal that parental shaming, temperamental inhibition, 524 Merrill-Palmer Quarterly and gender interact to influence shame responding in early childhood. This is a novel contribution pointing to several important sources of individual differences in early childhood shame responding. Moreover, it appears that both mothers and fathers play a significant role in the development of shame both for girls and for boys. Second, the data are the first to indicate that shame responding in early childhood predicts higher or lower proneness to shame in middle childhood, conditional on gender and parenting factors. As such, they suggest that early childhood shame responding may contribute to subsequent proneness to shame. Further research is needed to assess its longitudinal relations with maladaptive outcomes and the role shame proneness may play in such outcomes (Ferguson et al., 2000; Ferguson et al., 1999; Luby et al., 2009). References Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage. Alessandri, S. M., & Lewis, M. (1993). Parental evaluation and its relation to shame and pride in young children. Sex Roles, 29, 335–343. Alessandri, S. M., & Lewis, M. (1996). Differences in pride and shame in maltreated and nonmaltreated preschoolers. Child Development, 67, 1857–1869. Barrett, K. C. (1998). A functionalist perspective to the development of emotions. In M. F. Mascolo & S. Griffin (Eds.), What develops in emotional development? (pp. 109–133). New York: Plenum Press. Barrett, K. C., Zahn-Waxler, C., & Cole, P. M. (1993). Avoiders vs. amenders: Implications for the investigation of guilt and shame during toddlerhood? Cognition and Emotion, 7, 481–505. Belsky, J., Domitrovich, C., & Crnic, K. (1997). Temperament and parenting antecedents of individual differences in three-year-old boys’ pride and shame reactions. Child Development, 68, 456–466. Campos, J. J., Thein, S., & Owen, D. (2004). A Darwinian legacy to understanding human infancy: Emotional expressions as behavior regulators. Annals of the New York Academy of Sciences, 1000, 1–26. Chaplin, T. M., Cole, P. M., & Zahn-Waxler, C. (2005). Parental socialization of emotion expression: Gender differences and relations to child adjustment. Emotion, 5, 80–88. Cole, P. M., Barrett, K. C., & Zahn-Waxler, C. (1992). Emotion displays in twoyear-olds during mishaps. Child Development, 63, 314–324. Cole, P. M., Zahn-Waxler, C., & Smith, K. D. (1994). Expressive control during a disappointment: Variations related to preschoolers’ behavior problems. Developmental Psychology, 30, 835–846. Shame in Early and Middle Childhood 525 Crick, N. R., & Zahn-Waxler, C. (2003). The development of psychopathology in females and males: Current progress and future challenges. Development and Psychopathology, 15, 719–742. Denham, S. A. (1998). Emotional development in young children. New York: Guilford. Denham, S. A., Bassett, H. H., & Wyatt, T. (2007). The socialization of emotional competence. In J. E. Grusec & P. D. Hastings (Eds.), Handbook of socialization: Theory and research (pp. 614–637). New York: Guilford Press. Derryberry, D., & Rothbart, M. K. (1997). Reactive and effortful processes in the organization of temperament. Development and Psychopathology, 9, 633–652. Dickerson, S. S., Gruenewald, T. L., & Kemeny, M. E. (2004). When the social self is threatened: Shame, physiology, and health. Journal of Personality, 72, 1191–1216. Dickerson, S. S., & Kemeny, M. E. (2004). Acute stressors and cortisol responses: A theoretical integration and synthesis of laboratory research. Psychological Bulletin, 130, 355–391. Dienstbier, R. A., Hellman, D., Lehnhoff, J., Hillman, J. H., & Volkenaar, M. C. (1975). An emotion-attribution approach to moral behavior: Interfacing cognitive and avoidance theories of moral development. Psychological Review, 82, 299–315. Eisenberg, N., Cumberland, A., & Spinrad, T. L. (1998). Parental socialization of emotion. Psychological Inquiry, 9, 241–273. Ferguson, T. J., Stegge, H., Eyre, H. L., Vollmer, R., & Ashbaker, M. (2000). Context effects and the (mal)adaptive nature of guilt and shame in children [Monograph]. Genetic, Social, and General Psychology Monographs, 126, 319–345. Ferguson, T. J., Stegge, H., Miller, E. R., & Olsen, M. E. (1999). Guilt, shame, and symptoms in children. Developmental Psychology, 35, 347–357. Geppert, U. (1986). A coding system for analyzing behavioral expressions of selfevaluative emotions (Technical manual). Munich: Max Planck Institute for Psychological Research. Gershoff, E. T. (2002). Corporal punishment by parents and associated child behaviors and experiences: A meta-analytic and theoretical review. Psychological Bulletin, 128, 539–579. Hollingshead, A. B. (1975). Four factor index of social status. Unpublished manuscript, Department of Sociology, Yale University, New Haven, CT. Izard, C. E. (1979). The maximally discriminative facial movement coding system (Max). Unpublished manuscript, Office of Instructional Technology, University of Delaware, Newark, DE. Kagan, J., Reznick, J. S., & Snidman, N. (1987). The physiology and psychology of behavioral inhibition in children. Child Development, 58, 1459–1473. 526 Merrill-Palmer Quarterly Kelley, S. A., Brownell, C. A., & Campbell, S. B. (2000). Mastery motivation and self-evaluative affect in toddlers: Longitudinal relations with maternal behavior. Child Development, 71, 1061–1071. Keltner, D. (1995). Signs of appeasement: Evidence for the distinct displays of embarrassment, amusement, and shame. Journal of Personality and Social Psychology, 68, 441–454. Khaleque, A., & Rohner, R. P. (2002). Perceived parental acceptance-rejection and psychological adjustment: A meta-analysis of cross-cultural and intracultural studies. Journal of Marriage and Family, 64, 54–64. Kochanska, G. (1991). Socialization and temperament in the development of guilt and conscience. Child Development, 62, 1379–1392. Kochanska, G., DeVet, K., Goldman, M., Murray, K., & Putnam, S. P. (1994). Maternal reports of conscience development and temperament in young children. Child Development, 65, 852–868. Kochanska, G., Gross, J. N., Lin, M., & Nichols, K. E. (2002). Guilt in young children: Development, determinants, and relations with a broader system of standards. Child Development, 73, 461–482. Lagattuta, K. H., & Thompson, R. A. (2007). The development of self-conscious emotions: Cognitive processes and social influences. In J. L. Tracy, R. W. Robins, & J. P. Tangney (Eds.), The self-conscious emotions: Theory and research (pp. 91–113). New York: Guilford Press. Lewis, H. B. (1971). Shame and guilt in neurosis. New York: International Universities Press. Lewis, M. (1992). Shame: The exposed self. New York: Free Press. Lewis, M. (2007). Self-conscious emotional development. In J. L. Tracy, R. W. Robins, & J. P. Tangney (Eds.), The self-conscious emotions: Theory and research (pp. 134–149). New York: Guilford Press. Lewis, M., Alessandri, S. M., & Sullivan, M. W. (1992). Differences in shame and pride as a function of children’s gender and task difficulty. Child Development, 63, 630–638. Lewis, M., & Michalson, L. (1983). Children’s emotions and moods: Developmental theory and measurement. New York: Plenum Press. Lewis, M., & Ramsay, D. (2002). Cortisol response to embarrassment and shame. Child Development, 73, 1034–1045. Lindsey, E. W., & Mize, J. (2001). Interparental agreement, parent-child responsiveness, and children’s peer competence. Family Relations, 50, 348–354. Luby, J., Belden, A., Sullivan, J., Hayen, R., McCadney, A., & Spitznagel, E. (2009). Shame and guilt in preschool depression: Evidence for elevations in self-conscious emotions in depression as early as age 3. Journal of Child Psychology and Psychiatry, 50, 1156–1166. Shame in Early and Middle Childhood 527 Mills, R. S. L. (2003). Possible antecedents and developmental implications of shame in young girls. Infant and Child Development, 12, 329–349. Nelson, D. A., Hart, C. H., Yang, C., Olsen, J. A., & Jin, S. (2006). Aversive parenting in China: Associations with child physical and relational aggression. Child Development, 77, 554–572. Nix, R. L., Pinderhughes, E. E., Dodge, K. A., Bates, J. E., Pettit, G. S., & McFadyen-Ketchum, S. A. (1999). The relation between mothers’ hostile attribution tendencies and children’s externalizing behavior problems: The mediating role of mothers’ harsh discipline practices. Child Development, 70, 896–909. Paulhus, D. L., Robins, R. W., Trzesniewski, K. H., & Tracy, J. L. (2004). Two replicable suppressor situations in personality research. Multivariate Behavioral Research, 39, 303–328. Porter, C. L., Hart, C. H., Yang, C., Robinson, C. C., Olsen, S. F., Zeng, Q., et al. (2005). A comparative study of child temperament and parenting in Beijing, China and the western United States. International Journal of Behavioral Development, 29, 541–551. Robinson, C. C., Mandleco, B. L., Olsen, S. F., & Hart, C. H. (1995). Authoritative, authoritarian, and permissive parenting practices: Development of a new measure. Psychological Reports, 77, 819–830. Robinson, C. C., Mandleco, B. L., Olsen, S. F., & Hart, C. H. (2001). Parenting Styles and Dimensions Questionnaire (PSDQ). In J. Touliatos, B. F. Perlmutter, & M. A. Straus (Eds.), Handbook of family measurement techniques (3rd ed., pp. 188–189). Thousand Oaks, CA: Sage. Rothbart, M. K., Ahadi, S. A., & Hershey, K. L. (1994). Temperament and social behavior in childhood. Merrill-Palmer Quarterly, 40, 21–39. Rothbart, M. K., Ahadi, S. A., Hershey, K. L., & Fisher, P. (2001). Investigations of temperament at three to seven years: The Children’s Behavior Questionnaire. Child Development, 72, 1394–1408. Russell, A., Aloa, V., Feder, T., Glover, A., Miller, H., & Palmer, G. (1998). Sexbased differences in parenting styles in a sample with preschool children. Australian Journal of Psychology, 50, 89–99. Russell, A., Hart, C. H., Robinson, C. C., & Olsen, S. F. (2003). Children’s sociable and aggressive behavior with peers: A comparison of the US and Australia, and contributions of temperament and parenting styles. International Journal of Behavioral Development, 27, 74–86. Saarni, C. (1999). The development of emotional competence. New York: Guilford Press. Schore, A. N. (1996). The experience-dependent maturation of a regulatory system in the orbital prefrontal cortex and the origin of developmental psychopathology. Development and Psychopathology, 8, 59–87. 528 Merrill-Palmer Quarterly Stipek, D., Recchia, S., & McClintic, S. (1992). Self-evaluation in young children [Monograph]. Monographs of the Society for Research in Child Development, 57(1, Serial No. 226). Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics (4th ed.). Boston: Allyn and Bacon. Tangney, J. P. (1995). Shame and guilt in interpersonal relationships. In J. P. Tangney & K. W. Fischer (Eds.), Self-conscious emotions: The psychology of shame, guilt, embarrassment, and pride (pp. 114–139). New York: Guilford Press. Tangney, J. P., Burggraf, S. A., & Wagner, P. E. (1995). Shame-proneness, guiltproneness, and psychological symptoms. In J. P. Tangney & K. W. Fischer (Eds.), Self-conscious emotions: The psychology of shame, guilt, embarrassment, and pride (pp. 343–367). New York: Guilford Press. Tangney, J. P., & Dearing, R. L. (2002). Shame and guilt. New York: Guilford Press. Tangney, J. P., & Fischer, K. W. (Eds.). (1995). Self-conscious emotions: The psychology of shame, guilt, embarrassment, and pride. New York: Guilford Press. Tangney, J. P., Wagner, P. E., Burggraf, S. A., Gramzow, R., & Fletcher, C. (1990). The Test of Self-Conscious Affect for Children (TOSCA-C). Unpublished manuscript, George Mason University, Fairfax, VA. Tracy, J. L., & Matsumoto, D. (2008). The spontaneous expression of pride and shame: Evidence for biologically innate nonverbal displays. Proceedings of the National Academy of Sciences, USA 105, 11655–11660. Tracy, J. L., Robins, R. W., & Schriber, R. A. (2009). Development of a FACSverified set of basic and self-conscious emotion expressions. Emotion, 9, 554–559. Weiner, B. (1986). An attributional theory of motivation and emotion. New York: Springer-Verlag. Whisman, M. A., & McClelland, G. H. (2005). Designing, testing, and interpreting interactions and moderator effects in family research. Journal of Family Psychology, 19, 111–120. Winsler, A., Madigan, A. L., & Aquilino, S. A. (2005). Correspondence between maternal and paternal parenting styles in early childhood. Early Childhood Research Quarterly, 20, 1–12. Yang, C., Hart, C. H., Nelson, D. A., Porter, C. L., Olsen, S. F., Robinson, C. C., et al. (2004). Fathering in a Beijing, Chinese sample: Associations with boys’ and girls’ negative emotionality and aggression. In R. D. Day & M. E. Lamb (Eds.), Conceptualizing and measuring father involvement (pp. 185–215). Mahwah, NJ: Erlbaum.