Abstract | This article assesses the potential for reducing attrition bias by replacing survey dropouts with substitutes drawn from the same population and identified using propensity score matching. By linking register data with survey data, it is possible to observe unemployment outcomes for dropouts and therefore to test models of attrition. Doing so reveals the dropout process for unemployment to be nonignorable in this survey such that the commonly-used method of reweighting non-dropouts on the basis of sample frame information will be ineffective in overcoming attrition bias. The results indicate the effectiveness in theory of using substitutes but suggest that practical applications may only be successful where it is possible to incorporate information additional to that available in the sampling frame. Under such circumstances, it may similarly be possible to address nonignorable attrition by reweighting respondents. |
---|