The association between religiousness and depressive symptoms was examined with meta-analytic methods across 147 independent investigations (N ϭ 98,975). Across all studies, the correlation between religiousness and depressive symptoms was -.096, indicating that greater religiousness is mildly associated with fewer symptoms. The results were not moderated by gender, age, or ethnicity, but the religiousness-depression association was stronger in studies involving people who were undergoing stress due to recent life events. The results were also moderated by the type of measure of religiousness used in the study, with extrinsic religious orientation and negative religious coping (e.g., avoiding difficulties through religious activities, blaming God for difficulties) associated with higher levels of depressive symptoms, the opposite direction of the overall findings.

Both researchers and practitioners have demonstrated considerable interest in identity development, and positive personal identities have been associated with numerous positive mental health outcomes. However, major theories of identity development have neglected the salience of spiritual identity, even though the early work of William James placed this as a central component of personality. This article reviews four major theories of identity development (cognitive, psychodynamic, systems, and narrative) and suggests spiritual identity parallels to these theories. A tentative model of spiritual identity development is presented. Implications for therapy and future research concerning spiritual identity development are discussed.

Potential Moderators of the Association section) of results from Burris's (1994) previous study of intrinsic and extrinsic religious orientations and their associations with depressive symptoms was incorrect. Burris (1994) indeed found a positive association between extrinsic religious motivation and symptoms of depression. However, this effect was qualified by a significant Intrinsic ϫ Extrinsic interaction, such that persons who scored high on both orientations reported more symptoms of depression than did persons who scored high on intrinsic orientation only, whereas persons who scored low on intrinsic orientation reported a depression level that was between those two groups.Furthermore, the effect size (r) reported in Table 2 (p. 619) for this study should be .23, not Ϫ.11, and the sample size should be 200, not 100. These corrections result in a change in the reported random effects weighted average effect size (r) across all 147 studies (reported on p. 623, Omnibus Analysis section) from Ϫ.096 to Ϫ.094. All substantive conclusions of the review remain the same.