Testing the Impact of Mixed-Mode Designs (Mail and Web) and Multiple Contact Attempts within Mode (Mail or Web) on Clinician Survey Response

Objective. To compare response rate and nonresponse bias across two mixed-mode survey designs and two single-mode designs. Data Sources. This experiment was embedded in a clinician survey of knowledge and attitudes regarding HPV vaccination (n = 275). Study Design. Clinicians were randomly assigned one of two mixed-mode (mail/web or web/mail) or single-mode designs (mail-only/web-only). Differences in response rate and nonresponse bias were assessed. Principal Findings. Using a multiple-contact protocol increased response, and sending a web survey first provided the more rapid response. Overall, the mixed-mode survey designs generated final response rates approximately 10 percentage points higher than their single-mode counterparts, although only the final response differences between the mail-only and web/mail conditions attained statistical significance (32.1 percent vs. 48 percent, respectively; p = .005). Observed differences did not result in nonresponse bias. Conclusions. Results support mixing modes of survey administration and web-based data collection in a multiple contact survey data collection protocol. Key Words. Survey methods, mode effects, response rate, clinician surveys
Clinician surveys have long been used by health researchers and policy makers to assess clinician knowledge, attitudes, and behaviors, including practice patterns and guideline adherence. Unfortunately, response rates to clinician [...]