Research on the relationship between religious commitment and psychopamology has produced mixed findings. In a recent meta-analysis, Bergin (1983) found mat 23% of the studies reported a negative relationship, 47% reported a positive relationship, and 30% reported no relationship at all between religion and mental health. Based on our review of more than 200 studies, we have discovered four additional trends: (a) Most studies linking religious commitment to psychopathology have employed mental health measures that we have called “soft variables,” that is, paper-and-pencil personality tests which attempt to measure theoretical constructs. In contrast, most of the research linking religion to positive mental health is on “hard variables,” that is, “real life” behavioral events which can be reliably ob-served and measured and which are unambiguous in their significance. (b) Low levels of religiosity are most often associated with disorders related to undercontrol of impulses, whereas high levels of religiosity are most often associated with disorders of overcontrol. (c) Behavioral measures of religious participation are more powerfully associated with mental health than are attitudinal measures. (d) Distinctions such as that between intrinsic and extrinsic religiosity explain some inconsistent findings.

No abstract