Abstract
In every survey based on a probability sample the problem arises of missing information, in addition to problems of sampling error and response error. Entire interviews may be missing because of non-response (refusals, not-at-homes, etc.) or because dwelling units which should have been in the sample were missed. An interview may be taken with one designated respondent in a family but not with others. Also, interviews may be incomplete.
This article explains the techniques that can be employed to deal with missing information.
Get full access to this article
View all access options for this article.