Cognitive interviewing techniques are quickly gaining recognition as useful methods for pre­testing and designing questionnaires. Conducting cognitive interviews to test a questionnaire allows researchers to understand how respondents interpret a question and what mental processes they must engage in to answer the question. With this information, researchers can assess whether questions are being interpreted in the intended manner, whether respondents are able to answer certain questions depending on its cognitive demands, and so on. Depending on the time and funding available, cognitive interviews may either be used alone or as a predecessor to field tests or other pretesting methods. The advantage of using other pretesting methods such as field tests is that they allow the researcher to see how the questionnaire works in a standard interview format outside of a controlled laboratory setting. In addition, after conducting cognitive interviews and making changes to the question­naire based on these interviews, a field test provides some feedback on recommended changes.

Field tests are usually conducted either with a single, newly-revised questionnaire or as a split­panel experiment in which the revised question­naire is tested against the original version. However, there is little work to date that systematically investigates whether results from cognitive interviewing generalize to a field setting. This is particularly important for surveys which, due to time constraints, are only pretested in a laboratory setting. It could be that the observations made in a controlled laboratory environment using cognitive interviewing techniques are not applicable to a field setting since the response process of the two interviewing methods are very different.

This paper compares cognitive interviewing techniques with standard field interviewing techniques using a dietary intake questionnaire. The paper presents some background information addressing the initial cognitive research done using the dietary intake questionnaire, some discussion of cognitive versus standard interviews, and a description of the present research comparing the two interviewing methodologies. Finally, the implications of adopting recommendations from cognitive inter­views without first testing them in a field setting are discussed.