Why not find out whether religious beliefs predict surgical outcomes? If they do, why not find out why? Reply to Freedland (2004).

In this reply to K. E. Freedland's (see record 2004-13299-002) comments on R. J. Contrada et al. (see record 2004-13299-001), it is shown that the statistical issues he raised, and his preferred interpretation of the findings, were adequately addressed in the original article. It is argued that methodological limitations also were fully characterized and do not differ in kind from those of biomedical studies. Other issues discussed include the merits of focusing on distal versus proximal causation, plausibility of explanatory mechanisms for health effects of religious involvement, and potential practical applications that do not require manipulation of religious involvement. The article is concluded by commenting on subtle aspects of discourse that may unnecessarily polarize discussions of possible physical health effects of religious involvement.