An IRT Analysis of the Prodromal Questionnaire-Brief Child Version: Developing a Screening Form that Informs Understanding of Self-Reported Psychotic-Like Experiences in Childhood

The Prodromal Questionnaire-Brief Child Version (PQ-BC) has been developed as a tool for identifying psychotic-like experiences (PLEs) in school-age children. The current study examined the psychometric properties of the PQ-BC, examined how well the PQ-BC estimates the latent construct of PLEs (θ ̂), and began the process of developing a screening form informed by item response theory (IRT). Utilizing the baseline (n=11,129) sample from the Adolescent Brain Cognitive Brain (ABCD) study, we examined which PQ-BC items provide the most information and best discriminate individuals experiencing PLEs. Using hierarchical linear models (HLMs), we found that θ ̂ scores were significantly associated with several previously identified predictors of psychosis spectrum symptoms (i.e., history of psychosis, internalizing symptoms, cognitive impairments, developmental milestone delays, and resting-state functional connectivity impairments) at baseline and year 1 (n=5,532). Using item level information and discrimination parameters of the PQ-BC from the baseline sample, we created a seven-item screening form. HLMs generally found significant associations between screening form scores for both baseline and year 1 with the aforementioned predictors. The analyses provide evidence for the validity of a screening form derived from the PQ-BC using IRT derived parameters. This screening form could prove useful when the full measure is not feasible.