Date of Award


Level of Access Assigned by Author

Open-Access Dissertation

Degree Name

Doctor of Philosophy (PhD)




Peter J. LaFreniere

Second Committee Member

Janice Zeman

Third Committee Member

Cynthia A. Erdley


The present study examined the development of guilt and shame in preschool children, as well as individual differences related to the expression of these emotions. Sixty-one children in three age groups were videotaped in a mishap paradigm in which an experimentally manipulated doll's arm fell off during play. Children were randomly assigned to either an ambiguous or a personal responsibility condition. Videotapes were coded for behavioral (e.g., latency to repair, avoidance) and affective (e.g., joy, tensionlwony) reactions. Individual differences were assessed through parental reports using the My Child (Kochanska, DeVet, Goldman, Murray, & Putman, 1994) and teacher ratings using the Social Competence and Behavior Evaluation Inventory (SCBE; LaFreniere, Dumas, Capuano, & Dubeau, 1992). As expected, 4-year-olds expressed fewer avoidant behaviors such as toy avoidance, experimenter avoidance than did 2 or 3-year-olds. Additionally, older preschoolers expressed more guilt-relevant emotions such as sadness and decreased joy, rather than shame-relevant emotions such as tension/worry, which were seen in younger preschoolers. Results of the responsibility manipulation were contrary to hypotheses. Children in the personal responsibility condition expressed more shame-relevant emotions and behaviors than did children in the ambiguous responsibility condition. Results of the guilt and shame classification received only partial support, as children were dichotomized according to avoidant behaviors. Results suggest that the avoidance reflects shame-prone responding in children, as shame is conceptualized as avoidant behavior and affective discomfort. Nonavoiders may not fully reflect guilt, as the groups were dichotomized by avoidance but not by reparations. Results suggest that nonavoiders may reflect at least a child's proneness to guilt, as nonavoiders were higher than avoiders in guilt-relevant reactions such as latency to repair. Additionally, the avoidanthonavoidant classification was related to age. However, responsibility manipulations did not relate to the avoidant classification as expected. Individual differences were associated with the nonavoidantlavoidant classification, reflecting guilt and shame-prone responding. Nonavoiders, in comparison with avoiders, were rated by teachers as more socially competent, and by parents as higher in affective discomfort after wrongdoing. These findings may suggest that shame developmentally precedes guilt, but that these emotions also reflect important individual differences in social and emotional functioning .