Social and academic profiles of externalizing and internalizing groups: Risk factors for emotional and behavioral disorders

Three groups of third-grade students classified as either Externalizers (n = 30) Internalizers (n = 55) or Controls (n = 96) were contrasted on 19 mea sures in social and academic domains using multivariate analyses. Based on these analyses, a series of descriptive discriminant function analyses were undertaken to determine the best predictors of Externalizing and Internali zing group membership. Of 19 variables, those that best predicted Externali zing status compared to Controls were critical events, social skills, acade mic competence, peer acceptance, negative narrative comments, math, and rea ding, in that order. Variables that best predicted Internalizing status com pared to Controls were social skills, academic competence, math, critical e vents, reading, peer acceptance, and negative narrative comments. Variables that best separated students on Externalizing and Internalizing status wer e critical events, disciplinary referrals, social impact, academic self-con cept, school absences, peer acceptance, and academic competence. Cross-vali dated classification analyses correctly classified 97% of the Externalizing group and 89% of the Internalizing group compared to Controls. A classific ation analysis contrasting Externalizers and Internalizers resulted in corr ect classification rates of 63% and 84%, respectively. Contrary to much res earch in special education, both Externalizing and Internalizing groups app eared to have adequate academic self-concepts and social self-concepts. The se findings were interpreted in light of recent work on positive illusory b iases and the hypothesized relationship between unrealistically positive se lf-evaluations and psychological adjustment.