Definition

Sampling bias

Systematic error introduced when the people (or units) in your sample differ from the population you want to describe.

Convenience samples of students, self-selected survey respondents, patients from one clinic — each recruits systematically unusual people, and no sample size fixes a biased sampling mechanism. The classic cautionary tales (election polls, WEIRD psychology samples) are all sampling-bias stories.

Defenses: probability sampling where possible, quotas or weighting toward known population margins, explicit reporting of who was reachable and who responded, and claims scoped to the population the sample can actually speak for.