Web Survey Bibliography
We present data gathered from two randomly-selected samples of 2,000 physician members of a large medical-specialty organization. Sampled physicians were asked to complete one of two surveys that differed in length and content. Participants were first sent links to a web-based version of the survey. While the fields remained open, participants without an email address on file and nonresponders were mailed a paper survey a maximum of four times and were sent links to the web-based version of the same survey a maximum of three additional times. The second paper survey mailing contained a $2 bill, and the final email and paper surveys each included an offer for a $10 gift card from an outside vendor. The final response rates to the surveys were 52.3% and 51.2%. Previous research suggests that there are modest amounts of response bias in surveys of physicians (McFarlane, Omsted, Murphy, & Hill, 2007). To test this idea using our samples, we compare respondenders to nonresponders and internet to paper survey responders using demographic variables available on the sampling frame, such as sex, age, subspecialty, geographic region, primary professional activity, and primary employer. Previous research has also demonstrated that pre-paid monetary incentives are associated with higher response rates in samples of physicians (Kellerman & Herold, 2001). We examine the impact of the pre- and postpaid incentives on response rates to our surveys.
Conference Homepage (abstract)