Comparing Physicians' Responses to the First and Second Mailings of a Questionnaire

Abstract

Surveys are an important data collection method in evaluation research. Low response rates often occur, with follow-ups used to increase responses. However, surveys of homogeneous groups, such as physicians, about topics of concern to them may not require high response rates. We examined physician responses to a first and second mailing of a family practice questionnaire sent to a national sample of 308 recent residency graduates, 383 residency directors, and 319 residents. Among graduates, the first mailing produced a 63% response and the second another 17%; among directors, thefirstproduced 79% and the second another 10%; and among residents, the first gave 55% and the second another 21%. No differences between the two mailings existed by demographic variables, substantive questions ornonresponse to questions, with two exceptions. The absence of differences in physicians who responded to the follow-up questionnaire suggests that remailings to obtain higher response rates may be unnecessary in some surveys.