I can look at data on smokers and recent ex-smokers to identify what variables correlate with success in quitting, or relapse. I can adjust for covariates, or leave the analysis unadjusted. I did a spot of Googling and coudldn't find analogues findings from the psychology of religion at the level of different denominations, although there were studies about religious behaviour, non-specific to particular faiths. Has this work been done, and if so, what are the determinants? If the work hasn't been done, which datasets are best suited for that?