Social Perceptions of Religiosity: Dogmatism Tarnishes the Religious Halo

Abstract

Researchers have provided much evidence in support of a religious halo effect—the tendency for people to evaluate a religious target more favorably than a nonreligious counterpart. This experimental study (N = 361 U.S. adults) extended previous work beyond the dichotomy of religious versus nonreligious by varying a fictional target's degree of belief. Only religious participants exhibited a nonspecific pro-religious bias consistent with the religious halo effect, favoring religious over not religious and very over somewhat religious targets. However, indicators of dogmatic thinking (e.g., I am so sure I am right about the important things in life, there is no evidence that could convince me otherwise) degraded target evaluations by religious and nonreligious participants alike. Results suggest that religiosity's positive associations depend upon an assumption of strong but flexible belief and underscore the importance of studying social perceptions of religious fundamentalists.

Notes

¹This definition of dogmatism closely aligns with the features of religious fundamentalism as described by Kellstedt and Smidt (1991).

^aParticipants were allowed to choose more than one racial identification category.

²Using data from a previous investigation of social perceptions and dogmatism in which participants filled out the DOG scale while taking the perspective of a “religious fundamentalist” target (Swan & Heesacker, 2010), we selected the four items (two reverse-scored) with the highest factor loadings (.67–.76) on an unrotated principal component analysis.

³An exploratory factor analysis using principal axis factoring and promax rotation revealed a single factor structure, with only one semantic differential item (weak—strong) loading on a second factor.

⁴Due to heterogeneous variances between experimental conditions (Levene's test), F(5, 341) = 5.22, p < .001, we chose to report the Welch-corrected F ratio and to apply Games-Howell alpha-adjustment to all p values resulting from follow-up mean comparisons.

⁵Given (a) that our sample included a higher proportion of not religious participants than the general population of U.S. adults (Pew Research Center, 2012) and (b) that separate analyses with somewhat religious and very religious participants yielded similar results, we chose to collapse the somewhat religious and very religious categories of participant religiosity. Thus, participants were classified as either not religious or religious for this analysis.

*n = 201.

**n = 156.