Effects of different monetary incentives on the return rate of a national mail survey of physicians.

Abstract

Mail surveys of physicians have been characterized by lower response rates than general population surveys, raising concerns about nonresponse bias. Although monetary incentives have routinely been used to improve survey response among physicians, questions remain regarding how much of an incentive is most cost-effective. The present study seeks to further examine the effects of incentive size on response rates to a national mail survey of physicians. This study used a random sample of 873 physicians practicing in the United States; the response rate was 65% (n = 563). Respondents were randomly assigned to receive a $5, $10, or $20 cash incentive in the initial mailing. Except for the magnitude of the incentive, the procedures for each condition were identical, with each respondent receiving up to 3 follow-up mailings and 2 telephone calls. Overall response rates ranged from 60.3% for the $5 incentive category to 68.0% for the $10 incentive category. Differences in overall response rates across the incentive categories, however, were not significant. Higher levels of incentives also did not significantly reduce the number of mail and/or telephone interventions required to reach the target response rate of 60.0%. As expected, aggregate costs (excluding labor) were lowest for the $5 incentive group. Our findings suggest that changes in the magnitude of incentive do not automatically result in increases in survey response among physicians. Possible reasons for this lack of effect as well as alternatives to monetary incentives are addressed.