Is it the perceived stress scale (PSS) Undimimensional and invariant? A Bifactor analysis in Mexican adults

Juárez-García A.; Merino-Soto C.; Brito-Ortiz J.F.; Nava-Gómez M.E.; Monroy-Castillo A. · Current Psychology (2023) · DOI: 10.1007/s12144-021-02067-x

Resumen

Abstract The perceived stress scale (PSS) has been widely used as a unidimensional global measure of stress despite its classical bi-dimensional structure. However, a bifactor model seems to be the best structure alternative in terms of a possible method effect. The few available studies on PSS validity (in both the 14 and 10 item versions) that test the bifactor model do not consider its measurement invariance between occupational status and gender. The goal of this study was to test the bifactor structure and measurement invariance between occupational status and gender in the PSS 14 and 10 item versions. 1149 Mexican adults voluntarily answered the Spanish version of the PSS, 591 were students and 558 were health or education workers. Through a SEM methodology, several models were tested (unidimensional, two factors, bifactor, positive and negative wording effect, and random content). The results showed a bifactor model and a general factor with a subfactor for negative phrasing as the best for both versions (14–10 items) (CFI >.998, RMSEA<.037). Measurement invariance for sex and occupational status was found. The PSS-10 version showed slightly better indicators. In conclusion, PSS-10 and PSS-14 versions showed no evidence of gender or occupational status bias. A bifactor structure was the best model in both versions; however, these findings suggest a clear method effect in the PSS due to items with different directions (negative phrasing). Suggestions for future research and practical implications are discussed.

Instituciones: Universidad de San Martín de Porres