Abstract
This study examined the unique longitudinal effects linking academic competence, social competence, and internalizing symptoms from childhood to adulthood. A multimethod and multi-informant approach was used to assess psychopathology and competence in 205 participants during four developmental periods. Social competence in childhood had a cascading effect on internalizing symptoms in adolescence, whereas social and academic competence in emerging adulthood had dual cascading effects on internalizing in young adulthood. Results suggested a developmental cascade beginning with externalizing symptoms in childhood, which contributed to lower academic achievement in adolescence, which in turn influenced social competence in emerging adulthood and internalizing symptoms in young adulthood.
Acknowledgments
This article is based on data collected as part of the Project Competence longitudinal study, which has been supported through grants to Ann Masten, Auke Tellegen, and Norman Garmezy from the William T. Grant Foundation, the National Science Foundation (SBR-9729111), the National Institute of Mental Health (R01MH33222), and the University of Minnesota. Preparation of this article was supported by a Killam Postdoctoral Research Fellowship from the University of British Columbia to Dr. Obradović. We express our deep appreciation to the participants for their many contributions to this endeavor over more than 20 years.
Notes
Note: AI = Adolescent Interview; CBCL = Child Behavior Checklist; CI = Child Interview; CRS = Competence Ratings Scales; PI = Parent Interview; PIAT = Peabody Individual Achievement Test; Rater-P = rating based on parent information; Rater-S = rating based on target information; RCP = Revised Class Play; SCL-90–R = Symptom Checklist-90–Revised; SQ = Status Questionnaire; YABCL = Young Adult Behavior Checklist; YASR = Young Adult Self Report; YSR = Youth Self Report.
a Includes ratings based on informant in question.
1Given the central nature of social withdrawal to both social competence and internalizing problems, we also examined the effects of construct overlap by reanalyzing the best-fitting model (i.e., Model 4) after removing indicators of withdrawn behaviors from internalizing problems. This modified model had borderline overall fit (CFI = .910, TLI = .891, RMSEA = .053). One path became nonsignificant in this model: from Soc1 to Int2 (−.32 changed to −.23). Paths from Aca3 and Soc3 to Int4 also decreased slightly (to −.37 and −.41, respectively) but remained statistically significant.
Note: Final selected model is presented in boldface. c =weighting constant for computing the chi-square statistic using robust estimation method; cd =weighting constant for the difference between two chi-square values using robust estimation. CFI = comparative fit index; TLI = Tucker-Lewis index; RMSEA = root mean square error of approximation.
a Scaling correction produced negative chi-square difference; therefore, standard chi-square difference test is reported for this analysis.