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?