Measuring fluctuations in paranoia: Validity and psychometric properties of brief state versions of the Paranoia Checklist.

Research increasingly assesses momentary changes in paranoia in order to elucidate causal mechanisms. Observed or manipulated changes in postulated causal factors should result in fluctuations in state paranoid ideation. Previous studies often employed a state-adapted Paranoia Checklist (Freeman et al., 2005) to measure state paranoia. This study examined whether the Paranoia Checklist or subsets of its items are appropriate for this purpose. Thirteen studies (N=860) were subjected to meta-analyses of each Paranoia Checklist item. We selected items based on (1) whether they showed pre-to-post change in the expected direction and (2) whether this effect was larger in experimental vs. control conditions. All resulting item selections were cross-validated on a hold-out sample (n=1893). Finally, we explored how much variation in paranoia was captured by the state-adapted version in a brief ambulatory assessment study (N=32). A thirteen item State Paranoia Checklist as well as a five item and a three item Brief State Paranoia Checklist were extracted. Cross validation revealed better model fit and increased sensitivity to change. Multilevel analysis indicated 25-30% of the variance in the Brief State Paranoia Checklists to be due to intra-individual daily fluctuations in paranoia. Our analyses produced reliable and valid revised scales. Increases in change sensitivity indicate that future assessment of state paranoia in experimental and ambulatory assessment studies can be optimized by using the revised scales.