
You are reviewing customer data and want to know whether one segment appears in your sample more or less often than expected from a known benchmark population. You need a statistical way to tell whether the gap is likely real or just sampling noise.
How would you check whether a customer segment is over- or under-represented in your data?