Back to AI News
arxiv

AltRA-Test: A Statistical Test for Replacing Human Annotators with LLM Judges

Researchers propose AltRA-Test, an alternative annotator test for settings where items are annotated by different subsets of annotators rather than a shared panel. The method uses a bootstrap that resamples items, draws k human annotations per item, and substitutes one with an LLM annotation to form paired all-human and mixed human-LLM panels. A replacement is supported when the mixed panel's inter-annotator agreement is statistically non-inferior to the all-human panel's. The authors establish asymptotic level-alpha consistency and evaluate the test on 18 LLMs and 5 prompting techniques across 20 datasets spanning text, image, audio, and video, finding about 100 items often suffice.

AI Brief

It gives researchers and evaluation teams a concrete statistical procedure for deciding when an LLM judge can justifiably replace human annotators, including evidence on how many items are needed.

arXiv:2611.26073v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used as judges and graders to replace human annotators, raising the question of when such replacement is statistically justified. To support such decisions, the Alternative Annotator Test (Alt-Test) was introduced for dense settings where all items are annotated by the same annotators. However, in many large-scale annotation settings, items are annotated by different subsets of annotators. We therefore propose the Alternative Annotator Test with Resampled Annotations (AltRA-Test), a statistical test that does not require that all items are annotated by the same annotators. AltRA-Test uses a bootstrap that resamples items, draws k human annotations for each item, and substitutes one of them with an LLM annotation. This produces paired all-human and mixed human-LLM panels. A replacement is supported when the inter-annotator agreement (IAA) of the mixed panel is statistically non-inferior to the IAA of the all-human panel, i.e., when the LLM preserves the signal that the human annotators share. We establish that AltRA-Test is a consistent test of asymptotic level alpha. We then evaluate it through simulations and real-world experiments in which we test 18 LLMs and 5 prompting techniques on 20 datasets across text, image, audio, and video. Across these evaluations, about 100 items often suffice for the test. In simulations, where we know whether the LLM is a better or a worse annotator than the humans, AltRA-Test makes more correct replacement decisions than existing baselines. In real-world data experiments, LLMs pass the test in about half of the setups, depending mainly on the task. Since replacement is not always justified, we encourage researchers to test it for their task.
Be the first to rate this article.

Comments & Questions (0)

Captcha
Please be respectful — let's keep the conversation friendly.

No comments yet

Be the first to comment!