Re-evaluation of the Latent Structure of Common Childhood Disorders: Is There a General Psychopathology Factor (P-Factor)?

Abstract

In the field of psychopathology, there is high comorbidity between different disorders. Traditionally, support for two broad correlated dimensions of internalizing and externalizing symptoms has consistently emerged for children and adolescents. To date, oblique 2 and 3 first-order factor models (factors for externalizing and internalizing, and fear, distress, and externalizing) and bi-factor models with the corresponding two and three group factors have been suggested for common internalizing and eternalizing child and adolescent disorders. The present study used confirmatory factor analyses to examine the relative support for these models in adolescents (≥ 12 to 18 years; N = 866) and children (6 to < 12 years; N = 1233) and the reliability and convergent and divergent validities of the psychopathology factor (P-factor) and group factors in the optimum bi-factor model. All participants were from a clinic and underwent Diagnostic and Statistical Manual of Mental Disorders, 4th Edition clinical diagnosis. The findings showed that the bi-factor model with two group factors (internalizing and externalizing) was the optimum model for both children and adolescents. For both groups, findings showed relatively higher reliability for the P-factor than the group factors, although the externalizing group factor showed substantial reliability in adolescents, and both the externalizing and internalizing group factors also showed substantial reliability in children. The factors of the optimum bi-factor model also showed good convergent and discriminant validities. The implications for theory and clinical and research practice related to psychopathology are discussed.