Linkage of different data sources is an intermediate step in many statistical processes. When dealing with data resulting from a record linkage process, it should be considered that the linkage is affected by two types of errors: false links and missed matches. If the linkage errors are not properly taken into account, i.e. standard statistical procedures are applied to the linked data, biased estimates and mis-relationships between variables recorded in different sources may result. This paper provides a sensitivity analysis of the effect of linkage errors on the estimation of linear and logistic regressions. Different linkage scenarios are proposed, with various matching variables and accordingly different linkage error levels. The analysis confirms the importance of linkage errors and highlights the relevance of missed matches. The effectiveness of the proposed adjustment methods is demonstrated even when the conditions for their applicability are not fully satisfied, however a framework for taking into account the complexity of linkage procedures is needed.
CITATION STYLE
Di Consiglio, L., & Tuoto, T. (2018). When adjusting for the bias due to linkage errors: A sensitivity analysis. Statistical Journal of the IAOS, 34(4), 589–595. https://doi.org/10.3233/SJI-170377
Mendeley helps you to discover research relevant for your work.