This is a revised proposal for a multidisciplinary study to investigate the psychosocial linkages between community relocation and health (mental and physical) for older women. Revisions made in response to IRG concerns include: Model - Indicators of the social environment are added; housing type is included as an exogenous variable; methods for controlling other sources of unobserved heterogeneity are specified. Measures - Secondary stressors and age density are specified; further assessment is reported for measures of self-assessments and relocation factors; issues of reliability are clarified. Data Analyses - Steps to build, estimate, and test the longitudinal model, including substantive hypotheses of the relocation- health process, are specified; issues in modelling multi-wave data are addressed. The sample will be 600 women, 55+ years, obtained from waiting lists of planned retirement communities (PRCs), naturally-occurring retirement communities (NORCs), and subsidized housing (SHs) in Madison and Milwaukee, Wisconsin. Sampling from these settings ensures variability in SES, age density, services provided, and other factors that are naturally-occurring interventions in relocation. The women will be interviewed within 2 months prior to moving, and 1 month, 6 months, and 1 year after moving. Relocation variables include """"""""push"""""""" (e.g., health problems) and """"""""pull"""""""" (e.g., location, services) factors that influence the decision to move; age density of new setting; and the voluntariness of the move. Secondary stressors include measures of chronic difficulties (e.g., increased isolation) that arise after relocation. Health status includes objective and subjective measures of physical health, depression (CES-D and NIMH- DIS), and positive psychological well-being. The psychosocial linkages, derived from self-concept theory, include the interpretive mechanisms (i.e., self-assessments through social comparisons, reflected appraisals, self-perceptions, psychological centrality) derived from the surrounding social environment, that influence how older women make sense of their relocation experiences in ways that maintain, enhance, or diminish their health status. Major categories of hypotheses are that (a) women with positive relocation profiles (e.g., high push/pull """"""""fit"""""""") will show better health following relocation; (b) women who engage in positive social comparisons, reflected appraisals, and self-perceptions will show better health following relocation and that (c) these interpretive mechanisms, derived from the social environment, will mediate or buffer the effects of relocation factors and secondary stressors on health. Furthermore, (d) women who engage in positive self-assessments in life domains (e.g., family) that are central to their identities will show especially positive health profiles. Thus, (e) overall, relocation sets up a process whereby secondary stressors, the social environment, interpretive mechanisms, and health are related through a dynamic, reciprocal causal structure. A variety of multivariate statistical techniques (multiple regression, analysis of variance, LISREL VII) will be used.
