Increasing the sample size of a survey is often thought to increase the accuracy of the results. However, an analysis of big surveys on the uptake of COVID-19 vaccines shows that larger sample sizes do not protect against bias.