eduzhai > Applied Sciences > Computer Science >

Estimating population average treatment effects from experiments with noncompliance

  • KanKan
  • (0) Download
  • 20210425
  • Save

... pages left unread,continue reading

Document pages: 40 pages

Abstract: Randomized control trials (RCTs) are the gold standard for estimating causaleffects, but often use samples that are non-representative of the actualpopulation of interest. We propose a reweighting method for estimatingpopulation average treatment effects in settings with noncompliance.Simulations show the proposed compliance-adjusted population estimatoroutperforms its unadjusted counterpart when compliance is relatively low andcan be predicted by observed covariates. We apply the method to evaluate theeffect of Medicaid coverage on health care use for a target population ofadults who may benefit from expansions to the Medicaid program. We draw RCTdata from the Oregon Health Insurance Experiment, where less than one-third ofthose randomly selected to receive Medicaid benefits actually enrolled.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...