The survey's low response rate does not automatically vitiate its findings. It does, however, weaken any claim that the respondents represent the entire profession unless the researchers show that nonrespondents were unlikely to differ in relevant ways.