A test of reducing callbacks and not-at-home bias in personal interviews by weighing at-home respondents

Statistical bias resulting from omission of survey respondents who are not at home when called is a growing problem for market researchers. Since callbacks are an expensive means of reducing this bias, many organizations weigh answers from at-home respondents by a reciprocal of their (self-reported) probability of being at home. This method creates distortions in some cases, sufficient to make its use inadvisable.