The Distribution of Gender Differences in the Temperament and Social Competence of Preschoolers

Abstract

The literature has shown gender differences on many temperament and social competence (SC) measures, though there are gaps in understanding where differences lie and whether it varies by informant. This study investigated how temperament relates to SC and whether gender is a moderator. Rater source and the use of standardized versus raw scores and how they influence gender as a moderator was a main focus. Temperament was measured by the CBQ (Putnam & Rothbart, 2006) and the newly-developed CBQ, Teacher Form (CBQ-T; Teglasi, Schussler, & Gifford, under review). SC was measured by the SCBE (LaFreniere & Dumas, 1992), and all measures were administered to the parents and teachers of preschoolers (N = 113; M age = 57 months). For temperament, findings supported the fact that rater agreement is low and holds true for both genders. On the temperament scales on which parents significantly differed from teachers, parents tended to rate boys more favorably than teachers. The hypothesis that teacher ratings would yield more gender differences than parents was supported. There were also more differences in variability between genders for teacher ratings, revealing that teachers tended to provide more extreme ratings. Scales with distributional differences were ones that have consistently yielded gender differences. For SC, girls had significantly higher means on several scales with raw scores, and raw scores produced more temperament x gender interactions for parent ratings. Activity Level and Anger had opposite effects for gender with higher activity predicting higher SC for girls and lower SC for boys, and higher anger predicting higher SC for boys and lower SC for girls. On Sadness, there were opposite rater effects with Sadness positively associated with SC for parents and negatively associated for teachers. There were fewer gender differences for teachers when considering correlations of temperament and SC. Overall, findings support the importance of obtaining information from parents and teachers about children's temperament and SC. Moreover, these results suggest that raw SC scores are more useful than standard scores for studying relations between SC and temperament, particularly with parent raters. Shortcomings included a limited sample precluding full examination of distributional differences.