Key points are not available for this paper at this time.
The desirability—and sometimes the necessity—of using mail ballots under certain conditions has led to considerable experimentation with methods of detecting or compensating for bias in this type of survey. This article summarizes the literature on the subject, and examines several methods which have been proposed to increase the reliability of mail surveys. It then goes on to develop a method for detecting bias in mail returns by applying “random order’ tests.
Robert Ferber (Thu,) studied this question.