Constantinou, Matthew; (2019) The Bifactor Model of Psychopathology: Methodological Issues and Clinical Applications. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text M_Constantinou_Thesis_Final_04.09.19_no sign.pdf Download (4MB) | Preview

Abstract

For decades, clinicians have debated whether psychiatric problems should be ‘lumped’ into broad dimensions or ‘split’ into discrete entities. The bifactor model provides a potential solution to this debate by including both a general dimension of psychopathological severity known as the p factor, and specific dimensions reflecting specific problem areas such as internalizing and externalizing. This thesis evaluates the methodological properties and clinical utility of the bifactor model. Chapter 3 is a reliability review of bifactor studies and demonstrates that while self-report measures capture both general and specific domains of psychopathology, the total and subscale scores derived mainly reflect a general p factor. Chapter 4 investigates whether the general and specific psychopathology factors are products of response biases (i.e. tendencies in the way people fill out questionnaires), rather than variation in people’s experiences of psychiatric problems. Less than 4% of the variance in the general and specific psychopathology factors was explained by response biases, demonstrating their substantive validity. Chapter 5 analyzes clinical outcomes assessed over a psychosocial intervention for antisocial youth with a bifactor model and demonstrates more nuanced changes in disorder-specific factors after accounting for changes in the p factor (e.g., antisociality declines but anxiety increases over time). Similarly, Chapter 6 demonstrates the prognostic value of specific personality disorders for predicting depression outcomes assessed over an inpatient intervention only after accounting for the prognostic effect of a general personality disorder factor (e.g., borderline personality disorder predicts slower recovery). These findings demonstrate the substantive nature of the general and specific psychopathology factors, but also the difficulties in reliably measuring specific domains beyond general psychopathology. They also support the bifactor model’s utility in untangling clinically relevant effects that are otherwise masked by the shared variance among psychiatric problems.

Download activity - last month

Export as CSVJSONXML

Download activity - last 12 months

Export as CSVXMLJSON

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item