Behavioral Science Dictionary

Response bias

Methods & Evidence

Systematic tendencies in how people answer that distort what surveys actually measure.

What it means

Response bias is a family of systematic tendencies in the way respondents answer questions that pull their answers away from their true states, contaminating self-report data. It includes acquiescence (the yea-saying tendency to agree regardless of content), social desirability (presenting oneself favorably on sensitive topics), extreme and midpoint responding (favoring the ends or the safe middle of a scale), and demand characteristics (answering to fit what one thinks the researcher wants). Because these tendencies operate consistently rather than randomly, they are a form of systematic measurement error that no increase in sample size will remove and that can manufacture spurious correlations when the same bias inflates several measures — the common-method problem. Researchers blunt it through neutral and balanced wording, reverse-keyed items, forced-choice and indirect formats, anonymity, and validity scales. It matters because so much of behavioral science rests on questionnaires, and uncorrected response bias can masquerade as substantive findings.

Examples

On a survey about exercise, many respondents report working out more than they actually do, because the socially approved answer biases their self-report upward.

Staff rate their manager glowingly on a survey the manager will read; make the same survey genuinely anonymous and the team's scores drop sharply.

Some respondents tick 'agree' straight down a questionnaire, endorsing both 'I enjoy crowds' and 'I prefer being alone' — the pattern reveals the habit, not the person.

Where this comes up

← All 1001 terms