Construct validation of self-reported stress scales.

Abstract

The recently developed Perceived Stress Reactivity Scale (PSRS; Schlotz, Yim, Zoccola, Jansen, & Schulz, 2011) examines perceived stress reactivity with 6 different subscales. The authors of the current study build on initial validation work with this scale by evaluating patterns of the convergent and discriminant validity of scores on its 6 substantially correlated subscales. Examination of the relationships between the 6 PSRS subscales and other variables (personality, depressive affect, eudaimonic well-being, environmental demands, and aspects of work) indicated some differentiation between the different facets of stress reactivity. The PSRS scales also correlated highly with the Perceived Stress Scale (PSS). To account for all of these relationships, a structural regression model specifying a hierarchical factor model for the PSRS scales showed that some relationships (e.g., with neuroticism) were best modeled at the general stress factor level using a latent variable also defined by both the PSS and the PSRS subscales. The model also demonstrated lower level specific relationships that generated better fit than when all predictors were forced to relate to the PSRS only through the general stress factor. Results indicate that domain-specific aspects of stress can be differentiated, which ultimately could prove useful in research on mental health consequences of stress originating from different life domains.