Showing papers by "Donald B. Rubin published in 2008"

For objective causal inference, design trumps analysis

Abstract: For obtaining causal inferences that are objective, and therefore have the best chance of revealing scientific truths, carefully designed and executed randomized experiments are generally considered to be the gold standard. Observational studies, in contrast, are generally fraught with problems that compromise any claim for objectivity of the resulting causal inferences. The thesis here is that observational studies have to be carefully designed to approximate randomized experiments, in particular, without examining any final outcome data. Often a candidate data set will have to be rejected as inadequate because of lack of data on key covariates, or because of lack of overlap in the distributions of key covariates between treatment and control groups, often revealed by careful propensity score analyses. Sometimes the template for the approximating randomized experiment will have to be altered, and the use of principal stratification can be helpful in doing this. These issues are discussed and illustrated using the framework of potential outcomes to define causal effects, which greatly clarifies critical issues.

For objective causal inference, design trumps analysis

Abstract: For obtaining causal inferences that are objective, and therefore have the best chance of revealing scientific truths, carefully designed and executed randomized experiments are generally considered to be the gold standard. Observational studies, in contrast, are generally fraught with problems that compromise any claim for objectivity of the resulting causal inferences. The thesis here is that observational studies have to be carefully designed to approximate randomized experiments, in particular, without examining any final outcome data. Often a candidate data set will have to be rejected as inadequate because of lack of data on key covariates, or because of lack of overlap in the distributions of key covariates between treatment and control groups, often revealed by careful propensity score analyses. Sometimes the template for the approximating randomized experiment will have to be altered, and the use of principal stratification can be helpful in doing this. These issues are discussed and illustrated using the framework of potential outcomes to define causal effects, which greatly clarifies critical issues.

Principal Stratification for Causal Inference With Extended Partial Compliance

Abstract: Many double-blind placebo-controlled randomized experiments with active drugs suffer from complications beyond simple noncompliance. First, the compliance with assigned dose is often partial, with ...

Matching with multiple control groups with adjustment for group differences

Abstract: When estimating causal effects from observational data, it is desirable to approximate a randomized experiment as closely as possible. This goal can often be achieved by choosing a subsample from the original control group that matches the treatment group on the distribution of the observed covariates. However, sometimes the original control group cannot provide adequate matches for the treatment group. This article presents a method to obtain matches from multiple control groups. In addition to adjusting for differences in observed covariates between the groups, the method adjusts for a group effect that distinguishes between the control groups. This group effect captures the additional otherwise unobserved differences between the control groups, beyond that accounted for by the observed covariates. The method is illustrated and evaluated using data from an evaluation of a school drop-out prevention program that uses matches from both local and nonlocal control groups.

Evaluating the effects of job training programs on wages through principal stratification

Abstract: In an evaluation of a job training program, the causal effects of the program on wages are often of more interest to economists than the program's effects on employment or on income. The reason is that the effects on wages reflect the increase in human capital due to the training program, whereas the effects on total earnings or income may be simply reflecting the increased likelihood of employment without any effect on wage rates. Estimating the effects of training programs on wages is complicated by the fact that, even in a randomized experiment, wages are truncated by nonemployment, i.e., are only observed and well-defined for individuals who are employed. We present a principal stratification approach applied to a randomized social experiment that classifies participants into four latent groups according to whether they would be employed or not under treatment and control, and argue that the average treatment effect on wages is only clearly defined for those who would be employed whether they were trained or not. We summarize large sample bounds for this average treatment effect, and propose and derive a Bayesian analysis and the associated Bayesian Markov Chain Monte Carlo computational algorithm. Moreover, we illustrate the application of new code checking tools to our Bayesian analysis to detect possible coding errors. Finally, we demonstrate our Bayesian analysis using simulated data.

Comment: The Design and Analysis of Gold Standard Randomized Experiments

Abstract: (2008). Comment: The Design and Analysis of Gold Standard Randomized Experiments. Journal of the American Statistical Association: Vol. 103, No. 484, pp. 1350-1353.

Outcome-free Design of Observational Studies: Peer Influence on Smoking

Abstract: For estimating causal effects of treatments, randomized experiments are appropriately considered the gold standard, although they are often infeasible for a variety of reasons. Nevertheless, nonran ...

Radiologist-oncologist workflow analysis in application of RECIST

Abstract: 6639 Background: Objective criteria for measuring response to cancer treatment are critical to clinical research and practice. Many trials use imaging modalities like CT and NCI's Response Evaluati...

Superconducting cavities for a B‐factory—Interim progress report

Abstract: Superconducting resonant cavities can economically provide the voltages needed to achieve the short bunches and high luminosity required for the e−e+ colliders proposed to produce B mesons. In this article a report is given of the progress in the development of a superconducting RF system for a B‐factory. (AIP)