Causal Effect Estimation and Inference Using Stata. (December 2017)

• Record Type: Journal Article
• Authors: Terza, Joseph V.
• Is Part Of: Stata journal. Volume 17:Number 4(2017)
• Page Start: 939
• Page End: 961
• Publication Date: 2017-12
• Subjects: st0506 -- margins -- causal effect estimation -- causal inference; Statistics -- Periodicals; Statistics -- Computer programs -- Periodicals; 001.422
• DOI: 10.1177/1536867X1801700410
• Languages: English
• ISSNs: 1536-867X

Abstract

Terza (2016b, Health Services Research 51: 1109–1113) gives the correct generic expression for the asymptotic standard errors of statistics formed as sample means of nonlinear data transformations. In this article, I assess the performance of the Statamargins command as a relatively simple alternative for calculating such standard errors. I note thatmargins is not available for all packaged nonlinear regression commands in Stata and cannot be implemented in conjunction with user-defined-and-coded nonlinear estimation protocols that do not make apredict command available. Whenmargins is available, however, I establish (using a real-data example) that it produces standard errors that are asymptotically equivalent to those obtained from the formulations in Terza (2016b) and the appendix available with this article. This result favors usingmargins (with its relative coding simplicity) when available. In all other cases, use Mata to code the standard-error formulations in Terza (2016b). I discuss examples, and I give corresponding Stata do-files in appendices.