Estimation of simple and correlated measurement variance when first and second interview use different survey modes

This paper discusses in detail issues dealing with the technical aspects of designing and conducting surveys. It is intended for an audience of survey methodologists.

This paper proposes a method for estimating simple and correlated measurement variance components when a re-interview is available for a subsample of respondents. However, the two measurements cannot be considered as being collected under the same conditions and, therefore, are subject to different measurement error variance. This consideration seems more realistic when, in actuality, it is impossible to ensure that the same measurement conditions are implemented in the two interviews, as in the case when operational and budget constraints suggest adopting a different survey mode for the second interview.