Modification, cross-validation, invariance, and latent mean structure of the self-presentation in exercise questionnaire

The Self-Presentation in Exercise Questionnaire (SPEQ; Conroy, Motl, & Hall, 2000) measures impression motivation (IM) and impression construction (IC) within the context of exercise. This study cross-validated 2 published models of SPEQ scores, evaluated the invariance of the SPEQ model between sexes, and tested for sex differences in latent means for IM and IC. Based on data from 711 adult exercisers, published models of SPEQ scores were found to be inadequate, and a specification search yielded a revised model comprising 2 correlated factors. This model demonstrated tight cross-validity in an independent sample of adult exercisers. Invariance of the scale between women and men also was supported. No sex difference existed in latent IM scores but men exhibited a higher IC latent mean than women. Priorities for future evaluations of the SPEQ include examining the invariance of the model in other populations and extending evidence for the validity of score interpretations.