Investigating Wording Versus Construct Effects for a Modified Perceived Stress Scale

• In Event: Exemplary Practices in Test Validity Research and Evaluation

Abstract

Two matched pairs of positively and negatively worded item sets were created based on the ten original items from the Perceived Stress Scale in an attempt to identify whether construct complexity or wording effects are the primary reason for the original two-factor structure that conflates these two factors. Results are discussed in the context of model fit as it relates to multidimensional constructs within a model, wording effects attributable to contrary interpretations, and wording effects attributed to contradictory interpretations with items impacted by issues such as re-centering.

Authors

• Ronna L. Turner, University of Arkansas

• Ki Matlock Cole, Oklahoma State University

• Wallace D. Gitchel, University of Arkansas at Little Rock