Causal Inference in Experiments with Mixed-Subjects Designs

Implements seven estimators for average treatment effect (ATE) estimation in mixed-subjects designs (MSDs), where human subjects data is augmented with predictions from large language models (LLMs). Includes Difference-in-Means, GREG, PPI++, Doubly-Tuned, Difference-in-Predictions (DiP), DiP++, and D-T DiP estimators. Provides point estimates, variance estimation via delta-method or bootstrap, and optimal design selection for budget allocation between human observations and LLM predictions.

mixedsubjects is a package for conducting social science experiments using the Mixed-Subjects Design and estimating causal effects. It implements seven estimators for average treatment effect (ATE) estimation in mixed-subjects designs (MSDs), where human subjects data is augmented with predictions from large language models (LLMs). Includes Difference-in-Means, GREG, PPI++, Doubly-Tuned, Difference-in-Predictions (DiP), DiP++, and D-T DiP estimators. Provides point estimates, variance estimation via delta-method or bootstrap, and optimal design selection for budget allocation between human observations and LLM predictions.

Installation

Interested users can install using:

# install.packages("remotes")
remotes::install_github("klintkanopka/mixedsubjects")