# Right Answer, Wrong Score: Uncovering the Inconsistencies of LLM Evaluation in Multiple-Choice Question Answering

arXiv: https://arxiv.org/abs/2503.14996 · HTML: https://mlsanity.com/paper/2503.14996/

## Key facts

- Cited by 28 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 25 references that are in the map.
- First posted on arXiv 19 Mar 2025.

## Builds on (most cited first, 17 of 17)

- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [Large Language Models are Zero-Shot Reasoners](https://mlsanity.com/paper/2205.11916/) (2022)
- [PIQA: Reasoning about Physical Commonsense in Natural Language](https://mlsanity.com/paper/1911.11641/) (2019)
- [Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering](https://mlsanity.com/paper/1809.02789/) (2018)
- [MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark](https://mlsanity.com/paper/2406.01574/) (2024)
- [CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge](https://mlsanity.com/paper/1811.00937/) (2018)
- [Large Language Models Are Not Robust Multiple Choice Selectors](https://mlsanity.com/paper/2309.03882/) (2023)
- [Leveraging Large Language Models for Multiple Choice Question Answering](https://mlsanity.com/paper/2210.12353/) (2022)
- [When Benchmarks are Targets: Revealing the Sensitivity of Large Language Model Leaderboards](https://mlsanity.com/paper/2402.01781/) (2024)
- [Artifacts or Abduction: How Do LLMs Answer Multiple-Choice Questions Without the Question?](https://mlsanity.com/paper/2402.12483/) (2024)
- [LLMs May Perform MCQA by Selecting the Least Incorrect Option](https://mlsanity.com/paper/2402.01349/) (2024)
- [What's the Meaning of Superhuman Performance in Today's NLU?](https://mlsanity.com/paper/2305.08414/) (2023)
- [Do Large Language Models Have an English Accent? Evaluating and Improving the Naturalness of Multilingual LLMs](https://mlsanity.com/paper/2410.15956/) (2024)
- [Truth or Mirage? Towards End-to-End Factuality Evaluation with LLM-Oasis](https://mlsanity.com/paper/2411.19655/) (2024)
- [ZEBRA: Zero-Shot Example-Based Retrieval Augmentation for Commonsense Question Answering](https://mlsanity.com/paper/2410.05077/) (2024)

## Built on by (most cited first, 28 of 28)

- [BMMR: A Large-Scale Bilingual Multimodal Multi-Discipline Reasoning Dataset](https://mlsanity.com/paper/2507.03483/) (2025)
- [Reasoning Models are Test Exploiters: Rethinking Multiple-Choice](https://mlsanity.com/paper/2507.15337/) (2025)
- [Improving LLM First-Token Predictions in Multiple-Choice Question Answering via Output Prefilling](https://mlsanity.com/paper/2505.15323/) (2025)
- [Movie Facts and Fibs (MF2): A Benchmark for Long Movie Understanding](https://mlsanity.com/paper/2506.06275/) (2025)
- [MMA-ASIA: A Multilingual and Multimodal Alignment Framework for Culturally-Grounded Evaluation](https://mlsanity.com/paper/2510.08608/) (2025)
- [On LLMs' Internal Representation of Code Correctness](https://mlsanity.com/paper/2512.07404/) (2025)
- [BenchMarker: An Education-Inspired Toolkit for Highlighting Flaws in Multiple-Choice Benchmarks](https://mlsanity.com/paper/2602.06221/) (2026)
- [Benchmarks Are Not Atomic: Composition-Aware LLM Evaluation using BenchHub](https://mlsanity.com/paper/2506.00482/) (2025)
- [Beyond MCQ: An Open-Ended Arabic Cultural QA Benchmark with Dialect Variants](https://mlsanity.com/paper/2510.24328/) (2025)
- [Revisiting Prompt Sensitivity in Large Language Models for Text Classification: The Role of Prompt Underspecification](https://mlsanity.com/paper/2602.04297/) (2026)
- [Improving Cross-Format Robustness in Language Models with Multi-Format Training](https://mlsanity.com/paper/2606.11643/) (2026)
- [C3PO: Evaluating Cross-Modal Composition and Counterfactual Performance in Omnimodal Models](https://mlsanity.com/paper/2608.05381/) (2026)
- [Quantifying LLM Biases Across Instruction Boundary in Mixed Question Forms](https://mlsanity.com/paper/2509.20278/) (2025)
- [You May Speak Freely: Improving the Fine-Grained Visual Recognition Capabilities of Multimodal Large Language Models with Answer Extraction](https://mlsanity.com/paper/2510.14885/) (2025)
- [Beyond Multiple Choice: Verifiable OpenQA for Robust Vision-Language RFT](https://mlsanity.com/paper/2511.17405/) (2025)
- [Unexplored flaws in multiple-choice VQA make benchmarking unreliable](https://mlsanity.com/paper/2511.22341/) (2025)
- [Exploring Fine-Tuning for In-Context Retrieval and Efficient KV-Caching in Long-Context Language Models](https://mlsanity.com/paper/2601.18527/) (2026)
- [Are We Evaluating Knowledge or Phrasing? Mitigating MCQA Sensitivity with ParaEval](https://mlsanity.com/paper/2606.10657/) (2026)
- [Punching Above Their Weight: Classification-Head Fine-Tuning of Tiny Language Models (TLMs) for Verifiable Multiple-Choice Tasks](https://mlsanity.com/paper/2607.03801/) (2026)
- [When Models Know When They Do Not Know: Calibration, Cascading, and Cleaning](https://mlsanity.com/paper/2601.07965/) (2026)
- [Intersectional Fairness in Large Language Models](https://mlsanity.com/paper/2604.20677/) (2026)
- [Impacts of Histories and Models on LLM Grading: A Study in Advanced Software Engineering Courses](https://mlsanity.com/paper/2606.08400/) (2026)
- [ForEx: A Formal Verification Framework for Explainable Reasoning in Logical Fallacy Detection and Annotation](https://mlsanity.com/paper/2606.21867/) (2026)
- [Enabling Cloud-Level Accuracy in Edge AI through IoT Data Preprocessing](https://mlsanity.com/paper/2606.22496/) (2026)
- [Benchmarking the Benchmarks: Testing the Predictive Validity of Commonsense Benchmarks](https://mlsanity.com/paper/2608.03340/) (2026)
- [When Models Defer to Wrong Answers: A Robustness Audit of Source-Attributed Cues in Multiple-Choice QA](https://mlsanity.com/paper/2609.08934/) (2026)
- [Likelihood Ranking doesn't Scale Like Prompting in LLMs](https://mlsanity.com/paper/2609.29390/) (2026)
- [RCCM: Diagnosing Gains and Harms in Reasoning-Prompt Transfer](https://mlsanity.com/paper/2610.19872/) (2026)

## Abstract (arXiv)

One of the most widely used tasks for evaluating Large Language Models (LLMs) is Multiple-Choice Question Answering (MCQA). While open-ended question answering tasks are more challenging to evaluate, MCQA tasks are, in principle, easier to assess, as the model's answer is thought to be simple to extract and is compared directly to a set of predefined choices. However, recent studies have started to question the reliability of MCQA evaluation, showing that multiple factors can significantly impact the reported performance of LLMs, especially when the model generates free-form text before selecting one of the answer choices. In this work, we shed light on the inconsistencies of MCQA evaluation strategies, which can lead to inaccurate and misleading model comparisons. We systematically analyze whether existing answer extraction methods are aligned with human judgment, and how they are influenced by answer constraints in the prompt across different domains. Our experiments demonstrate that traditional evaluation strategies often underestimate LLM capabilities, while LLM-based answer extractors are prone to systematic errors. Moreover, we reveal a fundamental trade-off between including format constraints in the prompt to simplify answer extraction and allowing models to generate free-form text to improve reasoning. Our findings call for standardized evaluation methodologies and highlight the need for more reliable and consistent MCQA evaluation practices.