Survey experiments in the study of religions combine survey methods and experimental methods to better understand causal relationships and generalize findings to a broader population. By using population-based sampling frames and randomly assigning people to experimental conditions, survey experiments can determine what causes certain patterns in religion. This approach helps address limitations of traditional surveys and experiments on convenience samples, allowing researchers to explore a wider range of questions and contexts in the social scientific study of religion.