“…It was not possible in this review to place studies in hierarchical order of methodological quality, due to the wide range of approaches used (qualitative and quantitative); lack of clarity over actual sample sizes especially where longitudinal surveys, and also nested case study approaches, were reported; differences in sizes, and types of samples (e.g., males only (Crosnoe & Elder, 2002;Vaillant & Mukamal, 2001), socially or educationally advantaged samples (e.g., Charbonneau-Lyons et al, 2002;Crosnoe & Elder, 2002;Palmore, 1979;Perls & Silver, 1999), occupation-specific cohorts and samples (Abraham & Hansson, 1995;Tate et al, 2003), samples biased toward healthier population members (Havighurst, 1957;Palmore, 1979;Williams & Wirths, 1965), members of organizations for older people (Everard et al, 2000), and in some cases lack of information about the area of the survey. Nor was it possible to compare and rank studies by the strength of their evidence, given that investigators defined successful aging largely in terms of their own disciplines, often using the concept uncritically, with the consequence that their findings reflected their definitions and measures in isolation from the broader literature.…”