Data Annotation with Large Language Models: Lessons from A Large Reanalysis Study

• In Event: Politics of AI Workshop III. The Impact of Large Language Models on Political Communication and Research

Brief Overview

We conduct a systematic replication and reanalysis of 14 recently published papers in top political science journals by re-annotating data that was previously coded by humans or supervised models with 15 different open-weight and proprietary LLMs.

Author

• Eddie Yang, Purdue University