Abstract
A previous study observed that reported childhood abuse moderated psychotic and emotional reactivity to stress among patients with non-affective psychotic disorder. However, that study used a type of analysis unsuited for skewed data. This study aimed (1) to replicate the study and (2) to examine whether we would obtain similar results using a statistical approach better suited to skewed data. Fifty-nine patients with non-affective psychotic disorder were examined for up to 6 days using an intensive diary method to assess levels of negative affect, psychosis, and daily-life stress. A mixed-linear regression largely replicated earlier findings, but a two-component analysis failed to replicate the moderating effect of reported childhood abuse. These results illustrate the importance of exploring different statistical approaches to skewed data. They may also indicate that stress sensitization does not offer a complete account for the effect of reported childhood abuse on psychotic symptom severity.