ABSTRACT
The often touted positive association between religion and wellbeing is mainly based on evidence from cross-sectional studies. This is problematic because such studies tend to draw conclusions at the individual level despite reporting associations at the group level. In addition to this fallacy, inferences at the group level are also likely to be inflated by the social desirability effect, which may further exacerbate misrepresentations of the individual level. To avoid these pitfalls, we examined prayer-wellbeing (P-WB) associations and social desirability effects at both levels, using single-level and multilevel regression analysis on a longitudinal dataset. Weekly reports of prayer and wellbeing from 282 frequently praying Danish Christians, totaling 4254 complete responses, were combined with a comprehensive background questionnaire featuring a social desirability measure targeting the religious domain.
A typical weak positive P-WB association was observed at the group level, which disappeared when controlling for social desirability. At the individual level, the average association across individuals was positive after controlling for social desirability. This overall relationship, however, conceals considerable individual variance with almost a fourth of the estimated individual P-WB associations going in the negative direction, emphasizing the need to be cautious when extrapolating group-level data to the individual level. These findings suggest that cross-sectional studies may oversimplify the P-WB relationship, especially, if the social desirability effect is ignored.
Acknowledgments
We wish to thank Lena Overgaard for helping recruit participants. Benjamin G. Purzycki and Uffe Schjoedt acknowledge support from the Consequences of Formal Education for Science and Religion Project grant that was funded by the John Templeton Foundation. Finally, we wish to thank the editor and the reviewers for several valuable comments and suggestions, which helped us improve the manuscript.
Disclosure statement
No potential conflict of interest was reported by the authors.
Supplementary material
Supplemental data for this article can be accessed on the publisher’s website.