Building a bridge between attachment and religious coping: tests of moderators and mediators

Abstract

Coping studies show that people often involve God in coping. Attachment studies suggest this should be particularly true for people with an insecure history, who use God as a surrogate figure to regulate distress (compensation hypothesis). The present questionnaire study is the first investigation of perceived attachment history and religious coping. It was tested if an insecure history is linked to involving God in coping, if parental religiousness moderates the association, and if religious coping mediates the link between an insecure history and compensatory religiosity characteristics. Results from 197 participants, from different religious and non-religious groups in the central parts of Sweden, supported the compensation hypothesis (i.e., an insecure history was linked to involving God). This support was moderated by parental religiousness. At low parental religiousness, religious coping mediated the association between an insecure history and compensatory religiosity characteristics. The Discussion proposes further cross-fertilization of attachment and religious coping.

Acknowledgment

Thanks to Lars Andersson and Staffan Wessberg for assistance with the data collections and to Berit Hagekull for generously setting aside time and for help with translations. The writing of this article was supported by a Grant (Dnr 1999-0507:01, 02) to Professor Hagekull from the bank of Sweden Tercentenary Foundation.

Notes

Assessments of romantic attachment (Bartholomew, & Horowitz, 1991) were also included. In line with previous studies showing romantic attachment to be less consistently linked to religiosity than attachment history (Granqvist, 2002b; Granqvist, & Hagekull, 2002), virtually no associations were obtained between romantic attachment and religiosity, regardless of whether attachment categories, prototype ratings, or a model-of-self and model-of-other (Kirkpatrick, 1998) distinction was used.

To avoid conflating within- with between-group variance, a dummy-coded sub-sample (0 = religious groups; 1 = remaining groups) was also included as a covariate in the analyses. The pattern of relations remained identical, and in most cases results were significant, but weaker, when sub-sample was partialled out. A strong relation between sub-sample and religiosity (e.g., Pearson r = 0.71 in relation to the religious coping factor score) was probably responsible for too little variance remaining on the religiosity variables for attachment to make a strong contribution.

Three out of the four correlations between religious coping and compensation and correspondence characteristics were significantly different, one (self-directing) at the conventional level (p < 0.05) and two (collaborative and the coping factor score) only at a marginal level (p < 0.10) of significance. These differences were all in the predicted direction (i.e., compensation being more strongly linked to involving God). Moreover, after controlling for general religious activity, the relation between the religious coping factor and the correspondence prototype disappeared, partial r = 0.04, n.s, whereas the link with the compensation prototype remained, partial r = 0.35, p < 0.001.

Mediation was tested and supported also for all three religious coping sub-scales.