Panel attrition
Panel attrition is the loss of members from a research panel or longitudinal study over time, as people stop responding, drop out or become unreachable. It shrinks the sample and can bias results if those who leave differ from those who stay.
How it works
Every panel loses members: people move, change email addresses, lose interest or tire of surveys. The risk is selective attrition: if younger, busier or less satisfied members leave faster, later waves over-represent the rest, and trends can reflect who remains rather than real change.
Researchers track retention by wave and compare leavers with stayers on known characteristics. They reduce attrition with reasonable survey frequency and length, fair incentives, reminders and updated contact details, and they refresh panels with new members. The American Association for Public Opinion Research notes that pushing attrition too low has its own risk: long-serving members can become conditioned.
A worked example
Quarterly surveys of 2,000 drivers make up an electric-vehicle charging network's satisfaction panel. After a year, 1,240 remain. Satisfaction appears to rise 6 points, but the members who left were less satisfied at the start. Weighting for that pattern shrinks the real rise to 2 points.
Related
Automate this in Pulse
Attrition in a long-running panel removes some kinds of member faster than others, and a Pulse interview with members who dropped out can ask what made them stop answering.