Interval Matching: Propensity Score Matching Using Case-Specific Bootstrap Confidence Intervals

• In Event: Advances in Propensity Score Analyses and Causal Inference

Abstract

The problem with propensity score (PS) matching is that it is difficult to establish a sensible criterion to evaluate the closeness of the matched cases without knowing the estimation errors of PS that are point estimates. Cochran and Rubin (1973) suggested using a caliper band in caliper matching to avoid “bad” matches. However, this case-invariant caliper band still cannot address this issue because the estimation errors of PS should be case-specific. The purpose of the present study is to extend caliper matching to a new PSM, interval matching, to accommodate the estimation error of PS for each case, using case-specific bootstrap confidence intervals. The implementation of interval matching is illustrated with an empirical example.

Author

• Wei Pan, Duke University