Because Atran [fixed link --ed.] mentioned him recently, I've looked up Jeremy Ginges, a social psychologist who does research relevant to conflict resolution. In his paper on Religion and Support for Suicide Bombing (PDF) he looks at the relation between religion and support for suicide bombing (as you might have guessed). Does religious devotion contribute to support for suicide bombing? His suggestion is both a yes and a no; particular beliefs don't seem to play much of a role, but there is a probability that religious rituals, insofar as they encourage expressions of devotion to the community, may play some role. Support for suicide bombing does not appear to be influenced by frequency of prayer among Muslims in Palestine -- the dominant form of religious devotion in Muslim life -- but it does appear to be influenced by mosque attendance. As he notes explicitly in the paper, this does not mean that mosque attendance in general increases support for suicide bombing, since that would obviously depend on things like local leadership and and ambient culture; what it does show is that even when there is an influence, there appears to be a sharp difference between religious acts like prayer and religious acts like mosque attendance, and it suggests that, to the extent religion is playing a role, it is doing so in the sense of providing a forum for leaders drumming up support for suicide bombing.
People who are interested in the work of Jeremy Ginges might also be interested in the work of Ara Naranzayan. I recommend his Yin and Yang and Heaven and Hell (PDF), which discusses the complex issues involved in studying the relationship between religion and intolerance. Naranzayan points to evidence that when you don't control for certain factors, measures of religiosity show some reliability in predicting intolerance, and when you control for those factors, measures of religiosity show some reliability in predicting tolerance. Prayer, for instance, appears to be consistently associated with tolerance; and, noteworthily, prayer is one of the features of religion most intensely associated with religious devotion and belief in God. So religious devotion appears to be a good predictor of at least certain kinds of tolerance; but the paradox is that religious devotion is also linked to certain predictors of certain kinds of intolerance. This explains why study of the subject has had such difficulty achieving any clear conclusions -- the relationship between religious factors and tolerance are not straightforward, but highly complex. He also notes more evidence along the lines noted in the Ginges paper linked above.
As always with this sort of thing, or indeed with most studies in cognitive science and social psychology generally, I recommend a grain of salt; the science on the topics studied is still usually in a very formative state -- there are all sorts of things that might not have been adequately studied yet that would shed new light on what scientists are studying. But a grain of salt is subtly different from skepticism; studies like the ones given here do represent a serious advance in our understanding of the issues, since they give us suggestive data useful for further inquiry.
