|
Abstract:
|
Deteriorating response rates and potential substantial nonresponse bias have been the concern for survey researchers. Nonresponse follow-up is not just a way to boost the response rate but also a means to learn our converted respondents, who otherwise would be the nonrespondents. This paper will study the nonresponse follow-up cases within different demographic sub-groups as to how they help reduce nonresponse bias for surveys with different response rates, complexities and topics. This study will utilize survey data collected from respondents from NORC's AmeriSpeak® Panel, a probability-based household panel with extensive effort on nonresponse follow-up. In the initial recruitment, sample units are invited to join AmeriSpeak via contacts by mail and telephone. The second-stage non-response campaign targets a random sub-sample of the nonrespondens from the initial recruitment. Units sampled for the non-response follow-up campaign are sent by express mail a new recruitment package with an enhanced incentive offer.NORC interviewers then make personal visits to the respondents' homes to encourage participation.
|
