It has become a commonplace to use response rates as a quality measure for surveys. This interpretation has conceptual challenges to it that are worth addressing, especially when looking at response rates over time, across different surveys, or across different modes of administration.

The talk attempts to motivate an alternative to the conventional interpretation of response rates. The perspective taken grows out of the employment of capture-recapture ideas to a reinterview program for both respondents and refusal conversion efforts for nonrespondents.

Usually, reinterview programs are aimed at content issues in surveys and have fallen into disuse as costs have risen and the belief in tighter survey controls from automation has become widespread. Efforts to address nonresponse have been focused on refusal conversion efforts and generally have not been successful in stemming the falloff in response in many surveys.

Come expecting to share in the ongoing discussion and to decide for yourself if there are practical alternatives to current practice.