Cognitive interviewing techniques are increasingly being used in questionnaire development to assess respondents' comprehension, their ability to retrieve information, and the processes they us in doing so. The results are then applied to construct new questions or to revise existing ones. Cognitive techniques have thus been used primarily at the "front end" of questionnaire development, prior to the pre- or field test. Because of the time needed to conduct and analyze such interviews, typically only a very small number of purposively- chosen respondents is interviewed at the questionnaire development phase. Recently researchers have begun to use "debriefing questions" following a standardized interview to assess respondents' reactions to the interview and, sometimes, to evaluate comprehension as well. This paper argues that debriefing questions can be used as a cognitive tool to assess the quality of the interview, and that because they can be administered to a large, representative sample, they are an important "back-end" supplement to the one-to-one interviews used in the early phases of questionnaire development. The paper (1) reviews the literature to see how respondent debriefing questions have been used in other surveys; (2) presents results of a respondent debriefing conducted in conjunction with the pretest of the Food Security Supplement to the Current Population Survey, and (3) recommends that such debriefings be made a routine part of questionnaire development and evaluation, along with such techniques as behavior coding.
