Research Reports

Can Cognitive Information be Collected through the Mail? Comparing Cognitive Data Collected in Written versus Verbal Form

Wendy Davis, Terry DeMaio, and Andy Zukerberg

KEY WORDS: Establishment surveys, Pretesting methods

ABSTRACT

Over the last decade, cognitive interviewing has become a primary tool in developing and pretesting questionnaires, even self-administered instruments. The presence of a cognitive interviewer may affect the respondent's strategy for completing a self-administered instrument. For example, the cognitive interview method requires that the form be completed in a contained period of time. This is especially relevant in the case of establishment surveys, since the forms may be very time-intensive to complete and the respondent often does not have the answers for all items and must consult colleagues or records.

This paper presents the results of a study comparing two different techniques for collecting cognitive information about the processes used by respondents to complete the Survey of Industrial Research and Development, an establishment survey sponsored by the National Science Foundation. One method was a concurrent think-aloud cognitive interview with probing. The second method was a mailout/mailback self-administered form with "cognitively-oriented debriefing" questions included directly on the form. The results compare the content of the cognitive information collected and the type of problems uncovered by each of the two methods.

Citation: 1995, Proceedings of the Section on Survey Research Methods, Alexandria, VA: American Statistical Association, pp. 906-910.