Validating the Factorial Structure of Ryff’s Scales of Psychological Well-being (18-items) Using Bayesian Structural Equation Modeling

• In Event: 349 - Section on Aging and the Life Course Roundtable Session
  In Refereed Roundtable: Table 09. Depression over the Life Course

Abstract

Springer and Hauser (2006) conducted confirmatory factor analyses on nationally representative data from three large surveys (MIDUS, NSFH II, and WLS) and discovered high correlations among four of the six Ryff’s Scales of Psychological Well-Being (SPWB) dimensions (personal growth, purpose in life, self-acceptance, and environmental mastery). Their findings questioned the discriminant validity of the SPWB and raised a serious concern about interpreting scores from these subscales. Using Bayesian Structural Equation Modeling (BSEM) approach, our paper aims to resolve the existing debate about the factorial structure of SPWB by analyzing the data derived from MIDUS wave one survey, which was also analyzed in Springer and Hauser’s study. The result showed that (a) the discriminate validity of Ryff’s SPWB was reasonable and a six-factor model was confirmed; (b) unfortunately, the factor loadings of few items were relatively small, which suggested the convergent validity of SPWB was not satisfied.

Authors

• Hsien-Yuan Hsu, National Taiwan Normal University

• TzeLi Hsu, Florida State University

• Sandra Acosta, Texas A&M University