PQ-B

Psychotic-like symptoms were assessed using the PQ-B (40), which consists of 21 items. The PQ-B is an abbreviation and refinement of the Prodromal Questionnaire [PQ-92; (40)], with added distress assessment, as distressing symptoms are thought to be more relevant in psychosis spectrum measurement. The items reflect symptoms and experiences that can appear in those meeting criteria for psychosis risk syndromes. Appropriate cut-offs vary widely by use scenario and desired sensitivity (35); among help-seeking outpatients Xu et al. (41) suggested cut-offs of 7 and 24 for the total and distress scores, respectively, to ensure sufficient sensitivity and specificity. As in the present study, the measure has also been used to examine PLEs in the general population (42–44).

Each PQ-B item is first rated based on whether one has ever experienced a symptom (yes / no), and then according to how much distress the symptom causes, on a five-point scale (strongly disagree / disagree / neutral / agree / strongly agree; coded 1– 5). The symptoms sum score is the number of endorsed symptoms and the distress score is the mean of the coded distress ratings (in calculating these scores missing values were substituted with intra-individual means).

For the current study, all paired PQ-B symptom and distress item responses were transformed into single categorical variables, with a “No symptoms” response as the lowest level below the distress response alternatives. A preliminary nominal factor modeling indicated that “Strongly Disagree” and “Disagree” responses were not distinguishable on a unidimensional latent scale, and these were therefore collapsed. Due to infrequent responses, the categories (starting from “Strongly Agree”) were collapsed as necessary into the previous one, for each group comparison separately, to ensure a minimum of five responses for each alternative in each group; the number of categories per item were thus 3– 5 (mean 4.0 across all analyses). The collapsing and final number of categories for the items in each analysis is presented in Supplemental Table 1.