Survey estimation under informative nonresponse with follow-up

This article deals with survey estimation when there is partial follow-up of sample nonresponse. Two different approaches that make use of the follow-up data are presented, the first based on weighting and the other on prediction, with appropriate variance estimators developed for each case. A simulation evaluation using synthetic data and informative nonresponse is then used to compare these two approaches, as well as to contrast them with a simpler weighting approach that ignores the information obtained by the follow-up survey and treats the nonresponse as missing at random. Our results indicate that the new approaches lead to significant improvement as far as estimation of the population total is concerned