MaCSim approach to assess the accuracy of individual matched records with varying block sizes and cut-off values
share
Record linkage is the process of matching together the records from different data sources that belong to the same entity. Record linkage is increasingly being used by statistical, health, government and business organisations to link administrative, survey, population census and other files to create a robust file for more complete and comprehensive analysis. Despite this increase, there has been little work on developing tools to assess the accuracy of linked files. Ensuring that the matched records in the combined file actually correspond to the same individual or entity is crucial for the validity of any analyses and inferences based on the combined data. Haque et al. (submitted for publication) proposed a Markov Chain based Monte Carlo simulation approach (MaCSim) for assessing linkage accuracy and used ABS (Australian Bureau of Statistics) synthetic data to illustrate the utility of the approach. Different blocking strategies were considered to classify matches from non-matches with different levels of accuracy. In order to assess average accuracy of linking, correctly linked proportions were investigated for each record. The analyses indicated strong performance of the proposed method of assessment of accuracy of the linkages. In this paper, this method is employed to evaluate the accuracy of linkages, using varying block sizes and different cut-off values while minimizing error. The aim is to facilitate optimal choice of block size and cut-off value to achieve high accuracy in terms of minimizing the average False Discovery Rate (FDR) and False Negative Rate (FNR).
READ FULL TEXT