The essential role of pair matching in cluster-randomized experiments, with application to the Mexican Universal Health Insurance Evaluation

  title={The essential role of pair matching in cluster-randomized experiments, with application to the Mexican Universal Health Insurance Evaluation},
  author={K. Imai and G. King and Clayton Nall},
  journal={Statistical Science},
A basic feature of many field experiments is that investigators are only able to randomize clusters of individuals — such as households, communities, firms, medical practices, schools, or classrooms — even when the individual is the unit of interest. To recoup some of the resulting eciency loss, many studies pair similar clusters and randomize treatment within pairs. Other studies (including almost all published political science field experiments) avoid pairing, in part because some prominent… Expand

Figures and Tables from this paper

Missing Data under the Matched-Pair Design : A Practical Guide ∗
The matched-pair design in experiments and matching in observational studies are powerful tools that enable robust causal inference. What if outcomes of some units are missing? First, deleting unitsExpand
Estimation of treatment effects in matched-pair cluster randomized trials by calibrating covariate imbalance between clusters.
We address estimation of intervention effects in experimental designs in which (a) interventions are assigned at the cluster level; (b) clusters are selected to form pairs, matched on observedExpand
Optimal allocation of subjects in a matched pair cluster-randomized trial with fixed number of heterogeneous clusters
A locally optimal design is proposed that accounts the clusters heterogeneity and optimally allocates the subjects within each cluster and to address the dependency of design on the unknown parameters is addressed. Expand
Why Match in Individually and Cluster Randomized Trials
This paper focuses on estimating the average treatment effect and uses the efficient influence curve to understand the information provided by each design for estimation of this causal parameter, and indicates that the pairmatched design is asymptotically less efficient than its non-matched counterpart. Expand
Optimality of Matched-Pair Designs in Randomized Controlled Trials
This paper studies the optimality of matched-pair designs in randomized controlled trials (RCTs). Matched-pair designs are examples of stratified randomization, in which the researcher partitions aExpand
Inference in Experiments with Matched Pairs
This paper studies inference for the average treatment effect in randomized controlled trials where treatment status is determined according to a “matched pairs” design. By a “matched pairs” design,Expand
Adaptive pair-matching in randomized trials with unbiased and efficient effect estimation.
The consequences of adaptive pair-matching in randomized trials for estimation of the average treatment effect, conditional the baseline covariates of the n study units are explored. Expand
A sensitivity analysis for missing outcomes due to truncation by death under the matched-pairs design.
This paper develops a sensitivity analysis for missing outcomes due to truncation by death and applies it to the matched-pairs design in observational studies and to the randomized evaluation of the Mexican universal health insurance program. Expand
Adaptive pre-specification in randomized trials with and without pair-matching.
A rigorous procedure to data-adaptively select the adjustment set, which maximizes the efficiency of the analysis and uses cross-validation to select from a pre-specified library the candidate targeted maximum likelihood estimator (TMLE) that minimizes the estimated variance. Expand
At What Level Should One Cluster Standard Errors in Paired Experiments, and in Stratified Experiments with Small Strata?
In paired experiments, the units included in the randomization, e.g. villages, are matched into pairs, and then one unit of each pair is randomly assigned to treatment. We conducted a survey ofExpand


A Politically Robust Experimental Design for Public Policy Evaluation, With Application to the Mexican Universal Health Insurance Program
  • G. King, E. Gakidou, +7 authors Hector Hernandez Llama
  • Medicine, Economics
  • Journal of policy analysis and management : [the journal of the Association for Public Policy Analysis and Management]
  • 2007
We develop an approach to conducting large-scale randomized public policy experiments intended to be more robust to the political interventions that have ruined some or all parts of many similarExpand
The merits of matching in community intervention trials: a cautionary tale.
It is shown in this paper that there are several analytic limitations with pair-matched designs, including the restriction of prediction models to cluster-level baseline risk factors, the inability to test for homogeneity of odds ratios, and difficulties in estimating the intracluster correlation coefficient. Expand
Simple sample size calculation for cluster-randomized trials.
This paper provides simple formulae for sample size determination for both unmatched and pair-matched trials in cluster-randomized trials, and illustrated using two case studies: an unmatched trial of the impact of impregnated bednets on child mortality in Kenya and a pair- matched trial of improved sexually-transmitted disease treatment services for HIV prevention in Tanzania. Expand
Variance identification and efficiency analysis in randomized experiments under the matched-pair design.
  • K. Imai
  • Mathematics, Medicine
  • Statistics in medicine
  • 2008
This paper extends Neyman's analysis to randomized experiments under the matched-pair design where experimental units are paired based on their pre-treatment characteristics and the randomization of treatment is subsequently conducted within each matched pair. Expand
Optimal multivariate matching before randomization.
The use of optimal multivariate matching prior to randomization to improve covariate balance for many variables at the same time is discussed, presenting an algorithm and a case-study of its performance. Expand
Clustered encouragement designs with individual noncompliance: bayesian inference with randomization, and application to advance directive forms.
Bayesian methodology for causal inference for the effect of the new target treatment versus the control target treatment in the randomized CED with all-or-none compliance at the unit level is proposed, which generalizes the approach of Hirano et al. (2000) in important and surprisingly subtle ways. Expand
Covariate-based constrained randomization of group-randomized trials
  • L. Moulton
  • Computer Science, Medicine
  • Clinical trials
  • 2004
This article describes a procedure – random selection from a list of acceptable allocations – to allocate treatment conditions in a way that ensures balance on relevant covariates, which has good frequency properties. Expand
What Do Randomized Studies of Housing Mobility Demonstrate?
A frame-work for causal inference when interference is present is developed and a number of causal estimands of interest are defined, which are of great importance for a researcher who fails to recognize that a treatment is beneficial when in fact it is universally harmful. Expand
On design considerations and randomization-based inference for community intervention trials.
This paper discusses design considerations and the role of randomization-based inference in randomized community intervention trials. We stress that longitudinal follow-up of cohorts withinExpand
Statistical analysis and optimal design for cluster randomized trials
In many intervention studies, therapy outcome evaluations, and educational field trials, random treatment assignment of clusters rather than persons is desirable for political feasibility, logistics,Expand