Validation of the Chinese version of the Perceived Stress Scale-10 integrating exploratory graph analysis and confirmatory factor analysis

Gen Hosp Psychiatry. 2023 Sep-Oct:84:194-202. doi: 10.1016/j.genhosppsych.2023.07.008. Epub 2023 Jul 31.

Abstract

Objective: The study aimed to initially assess the measurement properties of the 10-item simplified Chinese version of the Perceived Stress Scale (PSS-C-10) and as a first, assess a longitudinal measurement invariance (LMI).

Methods: A longitudinal survey was conducted with a convenient sample of healthcare students using the PSS-C-10. We assessed the PSS-C-10 mainly using composite analytical approaches, including exploratory graph analysis (EGA) and confirmatory factor analysis (CFA) to suggest the best-fit factor structure and assess measurement invariance.

Results: The EGA identified a two-factor structural solution with an accuracy of 98.6% at baseline and 100% at a 7-day follow-up. The CFA subsequently confirmed this structure, with a comparative fit index of 0.963 at baseline and 0.987 at follow-up, Tucker-Lewis index of 0.951 at baseline and 0.982 at follow-up, and root mean square error of approximation of 0.111 at baseline and 0.089 at follow-up. The LMI was supported by the goodness-of-fit indices, and their changes fell within the recommended cut-off range. Additionally, Cronbach's alpha (0.885 at baseline and 0.904 at follow-up), McDonald's omega (0.885 at baseline and 0.902 at follow-up), and an ICC value of 0.816 for 7 days demonstrated the robust reliability of the PSS-C-10.

Conclusion: The PSS-C-10 exhibited a stable two-factor structure with promising LMI and measurement properties.

Keywords: Confirmatory factor analysis; Exploratory graph analysis; Healthcare students; Longitudinal design; Perceived Stress Scale; Psychometrics.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Factor Analysis, Statistical
  • Humans
  • Psychometrics
  • Reproducibility of Results
  • Stress, Psychological* / diagnosis
  • Surveys and Questionnaires