Chi square test of independence
Do these differences by religious
affiliation reflect true differences in the
population?
Are the observed differences large
enough that we’re sure they’re not due
purely to chance?
Might we just have a weird sample?