Investigating response styles in self-report personality data via a joint structural equation mixture modeling of item responses and response times

Distinct, independent lines of experimental and model-based measurement research have converged on the presence of multiple, aberrant response styles potentially detrimental to the validity of self-report personality test scores. However, these two lines of research have been confounded with the nature in which the theoretical latent personality trait is measured – particularly, the item score and response time (RT)–which reflect structure and process, respectively, of the trait score. The purpose of the current study was to conjointly model item scores and RTs and extend this flexible, exploratory method beyond the domain of cognitive ability constructs to the detection and distinction between distinct latent (i.e., unobserved; unknown a priori) response styles in self-report personality data. Results supported multiple forms of distortion simultaneously present for two constructs, Dutifulness and Extraversion. Unintentional and intentional distortion dominated latent structures in former, whereas response sets dominated in the latter.