Influence functions for linear regression (with an application to regression adjustment)

DOI
Official URL
Abstract
Influence functions are useful, for example, because they provide an easy and flexible way to estimate standard errors. This paper contains a brief overview of influence functions in the context of linear regression and illustrates how their empirical counterparts can be computed in Stata, both for unweighted data and for weighted data. Influence functions for regression-adjustment estimators of average treatment effects are also covered.
Date Issued
2019
Publication Type
Working Paper
Subject(s)
Language(s)
en
Author(s)
Additional Credits
Publisher
University of Bern, Department of Social Sciences
Access(Rights)
open.access
Show full item