# Improving the Validity and Practical Usefulness of AI/ML Evaluations Using an Estimands Framework

arXiv: https://arxiv.org/abs/2406.10366 · HTML: https://mlsanity.com/paper/2406.10366/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 29 of its 66 references that are in the map.
- First posted on arXiv 14 Jun 2024.

## Builds on (most cited first, 29 of 29)

- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [TruthfulQA: Measuring How Models Mimic Human Falsehoods](https://mlsanity.com/paper/2109.07958/) (2021)
- [FaceNet: A Unified Embedding for Face Recognition and Clustering](https://mlsanity.com/paper/1503.03832/) (2015)
- [G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment](https://mlsanity.com/paper/2303.16634/) (2023)
- [Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models](https://mlsanity.com/paper/2206.04615/) (2022)
- [A Call for Clarity in Reporting BLEU Scores](https://mlsanity.com/paper/1804.08771/) (2018)
- [A Survey on Evaluation of Large Language Models](https://mlsanity.com/paper/2307.03109/) (2023)
- [Model Cards for Model Reporting](https://mlsanity.com/paper/1810.03993/) (2018)
- [State of What Art? A Call for Multi-Prompt LLM Evaluation](https://mlsanity.com/paper/2401.00595/) (2023)
- [Hidden Stratification Causes Clinically Meaningful Failures in Machine Learning for Medical Imaging](https://mlsanity.com/paper/1909.12475/) (2019)
- [Evaluating Large Language Models: A Comprehensive Survey](https://mlsanity.com/paper/2310.19736/) (2023)
- [Don't Make Your LLM an Evaluation Benchmark Cheater](https://mlsanity.com/paper/2311.01964/) (2023)
- [Are We Really Making Much Progress? A Worrying Analysis of Recent Neural Recommendation Approaches](https://mlsanity.com/paper/1907.06902/) (2019)
- [Lessons from the Trenches on Reproducible Evaluation of Language Models](https://mlsanity.com/paper/2405.14782/) (2024)
- [Data Cards: Purposeful and Transparent Dataset Documentation for Responsible AI](https://mlsanity.com/paper/2204.01075/) (2022)
- [The Benchmark Lottery](https://mlsanity.com/paper/2107.07002/) (2021)
- [Wider and Deeper LLM Networks are Fairer LLM Evaluators](https://mlsanity.com/paper/2308.01862/) (2023)
- [FOFO: A Benchmark to Evaluate LLMs' Format-Following Capability](https://mlsanity.com/paper/2402.18667/) (2024)
- [Cross-validation: what does it estimate and how well does it do it?](https://mlsanity.com/paper/2104.00673/) (2021)
- [Evaluation Gaps in Machine Learning Practice](https://mlsanity.com/paper/2205.05256/) (2022)
- [Benchmarks for Automated Commonsense Reasoning: A Survey](https://mlsanity.com/paper/2302.04752/) (2023)
- [Evaluating General-Purpose AI with Psychometrics](https://mlsanity.com/paper/2310.16379/) (2023)
- [LMentry: A Language Model Benchmark of Elementary Language Tasks](https://mlsanity.com/paper/2211.02069/) (2022)
- [Face Clustering: Representation and Pairwise Constraints](https://mlsanity.com/paper/1706.05067/) (2017)
- [Inferring Capabilities from Task Performance with Bayesian Triangulation](https://mlsanity.com/paper/2309.11975/) (2023)
- [Bridging the Gap between Reality and Ideality of Entity Matching: A Revisiting and Benchmark Re-Construction](https://mlsanity.com/paper/2205.05889/) (2022)
- [Between Progress and Potential Impact of AI: the Neglected Dimensions](https://mlsanity.com/paper/1806.00610/) (2018)
- [Estimating the Performance of Entity Resolution Algorithms: Lessons Learned Through PatentsView.org](https://mlsanity.com/paper/2210.01230/) (2022)
- [How to Evaluate Entity Resolution Systems: An Entity-Centric Framework with Application to Inventor Name Disambiguation](https://mlsanity.com/paper/2404.05622/) (2024)

## Built on by (most cited first, 6 of 6)

- [Agreement Metrics for LLM-as-Judge Evaluation: What to Report and Why](https://mlsanity.com/paper/2606.00093/) (2026)
- [When benchmark inferences do not compose: Projectibility in AI evaluation](https://mlsanity.com/paper/2607.26159/) (2026)
- [Frontier AI Forecasting Has a Measurement Problem: An Audit of Progress Evidence](https://mlsanity.com/paper/2608.14903/) (2026)
- [When 10,000 Windows Are Not 10,000 Tests: Auditing Statistical Confidence in Sliding-Window Time-Series Classification](https://mlsanity.com/paper/2609.30721/) (2026)
- [Are We Measuring Anticipation? Auditing Privileged Information in Procedural Video Evaluation](https://mlsanity.com/paper/2610.03826/) (2026)
- [Auditing Failure Attribution: Label Predictability and Judge Behavior](https://mlsanity.com/paper/2610.17747/) (2026)

## Abstract (arXiv)

Commonly, AI or machine learning (ML) models are evaluated on benchmark datasets. This practice supports innovative methodological research, but benchmark performance can be poorly correlated with performance in real-world applications -- a construct validity issue. To improve the validity and practical usefulness of evaluations, we propose using an estimands framework adapted from international clinical trials guidelines. This framework provides a systematic structure for inference and reporting in evaluations, emphasizing the importance of a well-defined estimation target. We illustrate our proposal on examples of commonly used evaluation methodologies - involving cross-validation, clustering evaluation, and LLM benchmarking - that can lead to incorrect rankings of competing models (rank reversals) with high probability, even when performance differences are large. We demonstrate how the estimands framework can help uncover underlying issues, their causes, and potential solutions. Ultimately, we believe this framework can improve the validity of evaluations through better-aligned inference, and help decision-makers and model users interpret reported results more effectively.