Unanswered prayers: religiosity and the god-serving bias.

Two self-report experiments examined how religiosity affects attributions made for a target person's death. Online adults (Study 1, N = 427) and undergraduate students (Study 2, N = 326) read about Chris who had a heart attack, used religious or health behaviors, and lived or died. Participants made attributions to Chris and God (both studies), and reported their emotions (Study 2). Participants made more attributions to Chris when he lived than when he died, but only when he used health behaviors. The highly religious made more attributions to God, but not when Chris used religious behaviors and died (the God-serving bias); they reported the most positive emotions when Chris lived after using religious behaviors (the Hallelujah effect). Directions for future research in terms of implicit religious beliefs and normative evaluations of religion are discussed.