Bias Arising in Wording Consumer Questionnaires

the same city. Consequently, the interview technique is frequently used to compare the consumption of various groupings of people within a community. If interview studies can be so designed that their biases are small, their data on consumption may be quite useful. In November 1951, the Economics Division of the Canada Department of Agriculture studied cheese consumption in Ottawa by both questionnaire and wholesale audit techniques. The apparent biases resulting from the questionnaire used were startling with respect to both quantity of cheese used and its frequency of use. The wholesale audit, after liberal upward adjustment for possible omissions, revealed that sales of cheese could not have exceeded 2.5 ounces per person per week while the consumer questionnaire indicated purchases of 3.4 ounces per person per week in the same area during the same month. Likewise, the consumer questionnaire indicated the very