An assessment of the construct validity of RyV’s Scales of Psychological Well-Being: Method, mode, and measurement eVects

This study assesses the measurement properties of RyV’s Scales of Psychological Well-Being (RPWB)—a widely used instrument designed to measure six dimensions of psychological well-being. Analyses of self-administered RPWB data from three major surveys—Midlife in the United States (MIDUS), National Survey of Families and Households II, and the Wisconsin Longitudinal Study (WLS)—yielded very high overlap among the dimensions. These large correlations persisted even after eliminating several methodological sources of confounding, including question wording, question order, and negative item-wording. However, in MIDUS pretest and WLS telephone administrations, The research reported herein was supported by the National Institute on Aging (R01 AG-9775 and P01 AG21079), by the William Vilas Estate Trust, and by the Graduate School of the University of Wisconsin-Madison. Computation was carried out using facilities of the Center for Demography and Ecology at the University of Wisconsin-Madison, which are supported by Center and Training Grants from the National Institute of Child Health and Human Development and the National Institute on Aging. We thank Corey Keyes for reading an early draft, sharing data, and code from RyV and Keyes (1995), and most of all for his collaborative spirit. We thank Carol RyV for oVering useful information about the construction of well-being measures at an early stage of this project. We thank Dirk van Dierendonck for sharing the factor correlations from his 2004 Personality and Individual DiVerences paper, Nora Cate SchaeVer for discussions of mode eVects, Tetyana Pudrovska for her keen editing of Table 2b, and Joe Savard for technical assistance. We also appreciate the constructive comments of two anonymous SSR reviewers. An earlier version of the paper was presented at the annual meetings of the Population Association of America, 2003. The opinions expressed herein are those of the authors. The WLS, MIDUS, and NSFH II data used in these analyses are publicly available at the following websites: http://www.ssc.wisc.edu/ wlsresearch/(WLS) and http://www.icpsr.umich.edu/access/index.html/ (MIDUS and NSFH). * Corresponding author. Fax: +1 608 265 5389. E-mail address: [email protected] (K.W. Springer). 0049-089X/$ see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.ssresearch.2005.07.004 K.W. Springer, R.M. Hauser / Social Science Research 35 (2006) 1080–1102 1081 correlations among the dimensions were much lower. Past research demonstrates that self-administered instruments provide more valid psychological measurements than telephone surveys, and we therefore place more weight on the consistent results from the self-administered items. In sum, there is strong evidence that RPWB does not have as many as six distinct dimensions, and researchers should be cautious in interpreting its subscales. © 2006 Elsevier Inc. All rights reserved.