Search

The article discusses the discrepancies arising out of communicative consumer questionnaires. To illustrate the point of how such bias can affect the results of a survey, the author presents a survey done by the Economics Division of the Canadian Department of Agriculture in 1951. The results of the survey differed from a wholesale audit by a large margin. The author performed two surveys of his own, based on the original, in an attempt to lower the discrepancy between the results of the survey and the figures found in the wholesale audit. His conclusions were that it is difficult to escape an upward bias in purchase data obtained by a consumer interview. Questions asking how often a product is used with no time-frame is more likely to produce erroneous results.