English title: Investigating the Effect of Bias in Survey Measures of Church Attendance

Author(s): Philip Brenner -

Language: English

Type: Journal article

Year: 2012

Abstract

That church attendance rates are overestimated by conventional surveys is well established. Much of the extant literature places the locus of the error primarily on measurement; overreporting on the part of the respondent. However, there has been relatively little research into the effect of this meas- urement error on the individual demographic predictors commonly associated with church attend- ance. In this paper, demographic subgroups are compared on their propensities to report church attendance in conventional surveys and time diaries across 14 countries and four decades. Findings indicate that these covariates are strongly correlated with both measures of attendance, but parame- ter estimates do not significantly or consistently differ between these modes. This finding suggests that, while conventional survey measures may overestimate population rates of attendance in some countries (i.e., North America), parameter estimates for these demographic predictors are largely unaffected by overreporting bias. Finally, limitations and future directions of research are discussed.

Volume: 73

Issue: 4

From page no: 361

To page no: 383

Refereed: Yes

DOI: 10.1093/SOCREL/SRS042

Journal: Sociology of religion