The religion-racism link reverses direction depending on the field conducting the study

Evidence

Field of Research. For studies performed or published in social science fields (k = 23), greater religiosity was associated with greater racism, fixed r = . (95% confidence interval [CI] = .18/.23); random r = . (95% CI = .13/.26). For studies performed or published in religious fields (k = 4), greater religiosity was associated with greater racial tolerance, although the effect was significant only under the fixed effects model, fixed r = –. (95% CI = –.21/–.16); random r = –. (95% CI = –.39/.20). — Deborah L. Hall and David C. Matz and Wendy Wood, Why Don’t We Practice What We Preach? A…, 66

Categorization

• XREF: The field-of-research moderating effect connects directly to the meta-analysis theme in Hall, Matz & Wood about self-reported conviction diverging from measured bias, and to the reader’s broader question about how research setting/sincerity shapes measured outcomes.
• SHARE: Would interest anyone tracking publication-bias or researcher-expectancy effects in psychology meta-analysis.

Related

• Conservative values partly mediate the link between religiosity and racism (similarity 0.81)
• Prior religion-prejudice meta-analyses were purely correlational and each studied one target group. (similarity 0.80)
• Self-reported religiosity reliably but weakly predicts racial prejudice (similarity 0.79)
• Correlational meta-analyses cannot establish causal links between religion and prejudice. (similarity 0.79)
• Religiosity-prejudice links are correlational, not established as causal (similarity 0.79)
• Religious conviction type rather than strength predicts measured racism (similarity 0.78)