Statistical Inference

Statistical inference is the set of methods for drawing conclusions about a whole population (母體) from a limited sample (樣本) of data, while honestly stating how uncertain those conclusions are. Instead of measuring every voter, patient, or customer, you measure a few and reason carefully about the rest.

Almost every claim you read — a poll says 52%, a drug lowers risk, version B converts better — is an inference, not a fact about the sample alone. The whole discipline exists because samples vary by luck: two random samples from the same population give different numbers, so inference quantifies that luck and stops you from over-reading noise.