Author: Rodney Stark (Baylor University)

  Download/Printing is only available to registered users. Please login.

ABSTRACT

If social science is to achieve valid universal theories, it is necessary to test them in as many different times and places as possible-hence the urgent need for more comparative research. To demonstrate this need, I review three recent instances wherein comparative research has revealed that (1) the proposition that religion functions to sustain the moral order is not universal, (2) most new religious movements are not the product of the discontent of the deprived but typically reflect the dissatisfactions of the privileged, and (3) the greater religiousness of women is not due to changes within Christianity but is a universal phenomenon. I then examine a set of pitfalls that often afflict quantitative comparative research that uses ecological or collective units of analysis such as nations or cities. Chief among these pitfalls are the ecological fallacy, cherry-picking of cases and variables, and the lack of comparability among cases. All three pitfalls are illustrated with recent examples.