A meta-analysis was conducted of the health/subjective well-being relationship, based upon studies of United States adults published prior to 1980. Dependent variables were zero-order and first-order correlations. A total of 231 zero-order effect size estimates were extracted from 104 different sources, while a total of 189 first-order effect sizes were extracted from thirty-three different sources. The majority of effect sizes were derived from samples where the midpoint of the age range was sixty-five and over. The 95 percent confidence interval estimates, generated using Tukey's jackknife procedure, were .29 to .35 and .29 to .38 for the zero-order and first-order correlations, respectively. Using weighted regression procedures, univariate analyses indicated that all but one study and sample characteristics were significant (p less than .05) predictors of zero-order and first-order correlations, though none of the predictors adequately specified variation in effect sizes. The mean value of the first-order correlations varied with the type of control variable, ranging from .42 (age) to .27 (SES composite). Health and subjective well-being were found to be positively and significantly related. Limitations of the meta-analysis were noted and suggestions were made for strengthening research on this topic.

No abstract

Meta-analytic techniques were used to synthesize findings from the extant empirical literature on gender and social class as predictors of subjective well-being among U.S. adults. Based upon Rubin's (1979) research, we tested a family of hypotheses that gender, relative to social class, accounted for more variation in subjective well-being. Examination of zero-order and first-order effect sizes revealed that men have slightly higher subjective well-being than women and that, while gender is a statistically significant predictor of subjective well-being, it accounts for little variance. Three measures of social class, composite socioeconomic status, occupational status, and income also are significantly related to subjective well-being on the zero-order level; but that relation is reduced when gender is used as a covariate. Overall, across the studies we synthesized, gender does not transcend social class as a major determinant of subjective well-being.

Meta-analytic techniques were used to synthesize findings on the social activity/subjective well-being relation. We chose zero-order and first order associations as our dependent variables and several different measure, sample, and study quality characteristics as our independent variables. We found that social activity is positively and significantly related to subjective well-being. Contrary to activity theory, informal activities and activities with friends were not related to subjective well-being consistently more strongly than were formal activities and activities with neighbors. In addition, the remaining measure, sample, and study quality characteristics were not adequate predictors of variation in activity/subjective well-being associations.

No abstract

No abstract

No abstract

No abstract