In search of the p-factor: An investigation using multi-informant data and higher-order factor modeling across measures of two personality disorder systems

Abstract

The Hierarchical Taxonomy of Psychopathology (HiTOP) includes a “p-factor”, representing covariation among all psychopathology dimensions. While some consider the p-factor a substantive dimension of psychopathology, others argue it may be an artifact resulting from single-rater biases. A similar debate exists in personality research, where the general factor of personality (GFP) has been shown to emerge only because of individual rater biases. Given the structural similarities between personality traits and psychopathology, single-rater assessments of psychopathology may be susceptible to such biases, potentially inflating covariance even more so than normal-range traits. This study examines the presence of the p-factor, using multi-informant data and higher-order factor modeling within two personality disorder frameworks: the Multi-source Assessment of Personality Pathology (MAPP) and the Personality Inventory for ICD-11 (PiCD). Em- ploying multitrait-multimethod confirmatory factor analysis (MTMM CFA) models and exploratory factor analysis (EFA), this study explores the structure of the psychopathology dimensions and the existence of the p-factor while addressing key limitations in the literature, including over-reliance on single-informant data and inadequate coverage of psy- psychopathology spectra. Results indicate that despite the fact that the higher-order factor structure may, to some degree, be exaggerated by the biases of individual raters, the p-factor does not seem to be entirely artifactual. Limitations, including the limited sampling of symptoms and the inability to make causal inferences based on latent variable modeling, are discussed.