Testing the Distinction Between Self-Serving and Self-Debasing Cognitive Distortions in Trait Aggression: A Multitrait-Multimethod Analysis of the CD-Quest.

Cognitive distortions are associated with aggression, but it remains unclear whether self-serving (SS) and self-debasing (SD) distortions represent distinct and criterion-relevant dimensions within the Cognitive Distortions Questionnaire (CD-Quest). This distinction is theoretically consequential because SS and SD distortions are thought to carry different implications for externalising versus internalising outcomes, and clarifying their separability has direct relevance for targeted assessment and formulation. This study examined whether SS and SD distortions are structurally distinguishable and whether they show differential associations with trait aggression after accounting for content overlap and rating-method variance. A community sample of 712 UK adults completed the CD-Quest and the Buss-Perry Aggression Questionnaire. Multitrait-multimethod confirmatory factor analysis was conducted using a correlated trait-correlated method minus one (CT-C(M-1)) parameterisation, with frequency ratings specified as the reference method and belief ratings modelled as the non-reference method. Structural models tested associations between latent SS, SD, belief-method factors, and aggression. The two-trait CT-C(M-1) model provided the best interpretable fit and significantly improved over the one-trait model, indicating that SS and SD were statistically distinguishable. However, their latent correlation was very high (r = .947), suggesting near-unidimensionality. Variance decomposition showed that CD-Quest responses primarily reflected substantive trait variance rather than rating-method effects. SS and SD showed comparable associations with aggression when examined separately, and constraining their paths to equality did not significantly worsen model fit. A general cognitive distortion factor was robustly associated with aggression, whereas the belief-method factor showed no incremental association. The CD-Quest supports a statistically detectable but weakly differentiated SS/SD structure. In community adults, aggression-relevant variance is better captured by a general cognitive distortion dimension than by separable SS- and SD-specific pathways.