Date of Award
MA in Clinical Psychology
The structure of psychopathology contributes to understanding the etiology and treatment of mental health disorders. Comorbidity is common, and the high correlation between dimensions may limit the research findings associated with a single dimension. The purpose of this study was to (1) evaluate different models of internalizing and externalizing in child psychopathology and (2) evaluate the relationships between other variables and dimensions of internalizing and externalizing across models that did and did not account for the correlation between dimensions. The first hypothesis was that a bifactor model, including a general psychopathology (P) factor and internalizing and externalizing factors, would provide the best model fit. The second hypothesis was that the relationships between variables and dimensions would differ across analyses. Baseline data from the parents of 294 clients ages 3 to 17 at a university associated community-based training clinic were used to test these hypotheses. The Youth Outcomes Questionnaire (Y-OQ) was used to indicate latent internalizing, externalizing, and general psychopathology (P) factors and Confirmatory Factor Analysis was used to evaluate a one-factor model, two-factor correlated model, and bifactor model. Age and gender variables from a demographic questionnaire and four scales (i.e., parenting efficacy, child difficulty, parenting consistency, and parental involvement in treatment) from the clinic’s Bimonthly Longitudinal Youth Questionnaire (BIL-Y) were used to test the second hypothesis.
The first hypothesis was supported, as the bifactor model provided the best fit; however, the internalizing items loaded more on the P factor than the internalizing dimension. The comparison of relationships between variables and internalizing and externalizing across regressions, a correlated two-factor model, and the bifactor model indicated that findings do differ across methods for all variables except parental involvement in treatment. These findings indicate that the interpretations one makes about variables and their relationship with internalizing and externalizing are dependent on if and how the correlation between internalizing and externalizing is addressed in the analysis. In conclusion, researchers must account for the correlation between internalizing and externalizing when studying the predictors and treatment of one dimension with either a bifactor model or a regression model that covaries for the other dimension.
Faber, Aubrey, "The confusing problem of overlapping internalizing and externalizing dimensions: What should we do?" (2020). Theses and Dissertations. 187.