Non-response bias
Non-response bias arises when non-respondents differ systematically from respondents on the variable of interest, distorting survey estimates. A larger sample does not fix it; remedies require follow-up, weighting, or design changes.
See it move
Of 500 staff surveyed, only 200 respond — and they are the most engaged and satisfied, so the observed mean satisfaction reads 7.8 out of 10. The 300 who stay silent are less engaged, and the true population mean is really 6.0 out of 10. Doubling the sample would not fix this: the bias comes from who answers, not from how many do.
Check yourself
A university places satisfaction questionnaires on dining tables during a Monday lunch service. Of 3,000 enrolled students, 400 forms are returned (a 13% response rate), and 78% of respondents rate the food as poor. A manager concludes that roughly 78% of all students are dissatisfied. What is the most significant flaw in this conclusion?