|
Edited on Sun Nov-30-08 03:24 AM by struggle4progress
First, there is a definitional problem: "religious faith" is a word, used to describe a variety of phenomena, that might include certain superstitious complexes, social conventions, ritual behaviors, ethical systems, ontological or teleological philosophies, and so on. Even if the associated "religion" is clearly identified, different "adherents" may "believe" quite differently and have entirely different motives. Thus, a particular "religious belief" may reflect a level of scientific knowledge, or gregariousness, or desire to conform, or enjoyment of ritual, or ethical presuppositions, or philosophical tendencies, or any of a number of things -- without designating any definite particular collection of personal preferences. A clear and meaningful definition is necessary
Second, there is the issue of cross-comparison between different societies. Crime statistics between different countries, for example, are known to be difficult to compare
Third, there is the question of identifying relevant variables, and appropriately reducing the collection of variables. One expects, for example, that crime levels may depend on socio-economic features; if so, one should correct for those variates, in order to investigate the expected effect of "faith" when "all else is equal." Absent a careful attempt to do so, one is left with possibilities such as both crime and expressions of religious faith are more common in poorer communities, but since poverty levels were not measured, the study finds crime correlated with faith
Fourth, there is the question of causality. Assuming that "religious faith" has been defined and measured in a meaningful way, and assuming that the cross-comparison is valid, and that proper corrections have been made for relevant variables, one should still ask: "Does an increase in the crime rate precede or follow an increase in expressed religious belief? Is there a coherent theory that explains the correlation?" and similar questions
|