Replication Materials for Balancing Large Language Model Alignment and Algorithmic Fidelity in Social Science Research

This repository contains code and data from the paper Balancing Large Language Model Alignment and Algorithmic Fidelity in Social Science Research

Steerability Prompts and Responses

This folder contains code and data for Study 1.

Pigeonholing Partisans Replication

Contains code and data for Study 2.

Human Ratings for Interrater Validation

Contains human and LLM data for validating LLM coding performance.

R Code and Data

Contains code and data for statistical analysis, figure creation. This code can be run end-to-end.

Online Appendix

PDF of the online appendix.