Zelig includes many specific methods, based on likelihood, frequentist, Bayesian, robust Bayesian, nonparametric, and population and superpopulation theories of inference. Zelig adds considerable infrastructure to improve the use of existing methods. It translates hard-to-interpret coefficients into meaningful quantities of interest, along with the uncertainty estimates (generalizing Clarify for Stata); automates graphics and numerical summaries for all models; evaluates counterfactuals (incorporating WhatIf); combines multiply imputed data sets to impute missing data and correct for measurement error (via Amelia); automates bootstrapping for all models; allows for matching for causal inference to reduce model dependence (via MatchIt and cem); and generates replication data files (for Dataverse to satisfy community replication standards). Developers can make their R packages usable from Zelig by writing a few simple bridge functions. If you have suggestions improvements, we’d love to hear from you!

Important Update!

Zelig is no longer being maintained, but a new package, {clarify}, is available to provide much of Zelig’s functionality for simulation-based inference of interpretable post-estimation quantities. Please see the {clarify} website for information on installing and using {clarify} and for performing tasks in Zelig workflow using {clarify}.