Premium
Estimation and Inference for the Causal Effect of Receiving Treatment on a Multinomial Outcome
Author(s) -
Cheng Jing
Publication year - 2009
Publication title -
biometrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.298
H-Index - 130
eISSN - 1541-0420
pISSN - 0006-341X
DOI - 10.1111/j.1541-0420.2008.01020.x
Subject(s) - test statistic , outcome (game theory) , causal inference , mathematics , statistics , multinomial distribution , econometrics , null distribution , statistic , randomized controlled trial , likelihood ratio test , null hypothesis , inference , statistical hypothesis testing , medicine , computer science , artificial intelligence , mathematical economics
Summary This article considers the analysis of two‐arm randomized trials with noncompliance, which have a multinomial outcome. We first define the causal effect in these trials as some function of outcome distributions of compliers with and without treatment (e.g., the complier average causal effect, the measure of stochastic superiority of treatment over control for compliers), then estimate the causal effect with the likelihood method. Next, based on the likelihood‐ratio (LR) statistic, we test those functions of or the equality of the outcome distributions of compliers with and without treatment. Although the corresponding LR statistic follows a chi‐squared (χ 2 ) distribution asymptotically when the true values of parameters are in the interior of the parameter space under the null, its asymptotic distribution is not χ 2 when the true values of parameters are on the boundary of the parameter space under the null. Therefore, we propose a bootstrap/double bootstrap version of a LR test for the causal effect in these trials. The methods are illustrated by an analysis of data from a randomized trial of an encouragement intervention to improve adherence to prescribed depression treatments among depressed elderly patients in primary care practices.