A demonstration of social desirability bias in survey research suitable for moderate to large class sizes is described. The method used to reduce subjects' response bias involved a manipulation of instructions on a questionnaire. The standard instruction format requested subjects to respond to three questions on dental hygiene behavior and not to sign their names on the questionnaires. The modified instructions included additional information regarding the anonymity of the subjects' responses, the need for accurate information, and the role of the subject as a contributor of valuable information. The results from the demonstration are reported and topics for classroom discussion are suggested.