Randomized Trial of Use of Incentive to Increase the Response Rate to a Mailed Survey

Low response rates, especially among health -care professionals, are a common problem in mailed survey research. Wc conducted a randonUzed trial to examine th e:: effec ts of cash incentives on response ratcs. A total of 3,335 Chinese medicine practitioners were randomized to one o f two interventions accompanying a mailed survey no incentive (n= 1,667), and monetary incentives in two levels at HKS 20, and J-IKS 30 (n:;834 in each group) on receipt of the returned questionnaire. The response rates were higher among those offered incentives than those without (34.7% vs. 28.5%. X l :; 14.34, p<O.OO t). but no significant differences were found between incentives at I-JKS 20 and HKS 30 (X2:;0.16, 1'=0.69). Although offered incentives can increase response rates, no incentive was the most cost-effective, in terms of cost per respondent (HKS21.90 per rcspondem). l-iowever, this study focuses o n CMPs, the findings may not represent the respo nse rate fo r all health-care professions in Hong Kong. In fact, the larges t population in the heahh care system in Hong Kong is nurses. lberefore, we cannot conclude that a money enclosure would have been more effective for all health-care professionals. Also, further systematic study of the effects of different incentive strategies in other related research should