Question-induced cognitive biases in reports of dietary intake by college men and women.

In two experiments, investigated how variations in questionnaire structure influence respondents' reports of two aspects of dietary intake--the frequency with which various food items are eaten and the sizes of the portions that are eaten. In Experiment 1, approximately 400 subjects, prior to making a frequency judgment, were asked to think either about a specific occasion or about all the occasions on which they had eaten a particular food. The thoughts that preceded the frequency judgment influenced that judgment: Thinking of the range of occasions on which a food is consumed resulted in higher frequency estimates than thinking only of the most recent occasion. In Experiment 2, the same subjects made judgments about their typical portion sizes of several foods relative to described standards. For only one of eight foods were estimates properly and significantly affected by differences among the described standards. These results suggest that respondents are not particularly sensitive to portion-size definitions. We consider the implications of these phenomena for the development of a general theory of the cognitive processes that subserve health-survey responding.