Effects of QR codes in a political science probability-based web survey

In this article, we use a probability-based, 30-minute web survey on political behavior to test effects of QR codes in mail invitations. As expected, we find a strong increase in mobile phone use compared to computer use. However, there is no difference in response rates, sample composition, response speed or response quality. This suggests that part of the sample is switching from computer to mobile phone, with no device effect. Importantly, the use of a QR code, which may give the impression of a short survey, led mobile phone users to perceive the survey as too long. This mismatch between expectations and the actual survey experience may have reduced panel consent rates in the QR code condition. While QR codes may be useful for pushing people to mobile devices, this result highlights the need for caution when using them for longer surveys, especially when completion or panel recruitment is important.