Inferential Statistics

Inferential statistics use a sample to draw careful conclusions about a larger population you did not fully measure, always with an attached estimate of how uncertain those conclusions are. It is the disciplined art of going from “what I saw in my data” to “what is probably true in general.”

For instance, tasting a spoonful of soup to judge the whole pot is inference: the spoon is your sample, the pot is your population. The catch is that the leap can go wrong if the sample is not representative or is too small, so inferential methods always come with caveats like confidence intervals (信賴區間) or p-values (p值) that quantify the risk. The number you compute is an estimate, not a guaranteed truth.