An Extended General Location Model for Causal Inferences from Data Subject to Noncompliance and Missing Values
2004
Peng, Yahong | Little, Roderick J. A. | Raghunathan, Trivellore E.
Noncompliance is a common problem in experiments involving randomized assignment of treatments, and standard analyses based on intentionâtoâtreat or treatment received have limitations. An attractive alternative is to estimate the ComplierâAverage Causal Effect (CACE), which is the average treatment effect for the subpopulation of subjects who would comply under either treatment (Angrist, Imbens, and Rubin, 1996, Journal of American Statistical Association91, 444–472). We propose an extended general location model to estimate the CACE from data with noncompliance and missing data in the outcome and in baseline covariates. Models for both continuous and categorical outcomes and ignorable and latent ignorable (Frangakis and Rubin, 1999, Biometrika86, 365–379) missingâdata mechanisms are developed. Inferences for the models are based on the EM algorithm and Bayesian MCMC methods. We present results from simulations that investigate sensitivity to model assumptions and the influence of missingâdata mechanism. We also apply the method to the data from a job search intervention for unemployed workers.
Показать больше [+] Меньше [-]Ключевые слова АГРОВОК
Библиографическая информация
Эту запись предоставил National Agricultural Library