The Measurement of Psychological Well-Being among Aging Adults

Kristen W. Springer, University of Wisconsin at Madison
Robert M. Hauser, University of Wisconsin at Madison

This project examines the measurement properties of Ryff's model of Psychological Well-Being (PWB) —-- a widely-analyzed scale designed to measure six dimensions of PWB. We analyzed MIDUS, NSFHII and WLS self-adminstered data using polychoric correlations with weighted least squares estimation and found substantial overlap between dimensions —-- some correlations reached .970. We ruled out several possible methodological causes for these results including question wording, question order, and negative item bias. We also examined the WLS phone items, asked via an unfolding technique. We found smaller factor correlations, yet unexpected bimodal distributions. Given the anomalous distribution of the phone items and the consistent results across MIDUS, NSFHII and WLS using self-administered instruments, we believe the phone results are caused by a misuse of the unfolding technique and do not support the existence of a multidimensional model. In sum, we caution researchers against using this model to explore the multidimensionality of PWB.