@inproceedings{balepur-etal-2026-benchmarker,
    title = "{B}ench{M}arker: An Education-Inspired Toolkit for Highlighting Flaws in Multiple-Choice Benchmarks",
    author = "Balepur, Nishant  and
      Rajasekaran, Bhavya  and
      Oh, Hyunjin Jane  and
      Xie, Michael  and
      Desai, Atrey  and
      Gupta, Vipul  and
      Moore, Steven James  and
      Choi, Eunsol  and
      Rudinger, Rachel  and
      Boyd-Graber, Jordan Lee",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.719/",
    pages = "15793--15824",
    ISBN = "979-8-89176-390-6",
    abstract = "Multiple-choice question answering (MCQA) is standard in NLP, but benchmarks lack rigorous quality control. We present BenchMarker, an education-inspired toolkit using LLM judges to flag three common MCQ flaws: 1) contamination{---}items appearing exactly online; 2) shortcuts{---}cues in the choices that enable guessing; and 3) writing errors{---}structural/grammatical issues based on a 19-rule education rubric. We validate BenchMarker with human annotations, then run the tool to audit 12 benchmarks, revealing: 2) contaminated MCQs tend to inflate accuracy, while writing errors tend to lower it and change rankings beyond random; and 3) prior benchmark repairs address their targeted issues (i.e., lowering accuracy with LLM-written distractors), but inadvertently add new flaws (i.e. implausible distractors, many correct answers). Overall, flaws in MCQs degrade NLP evaluation, but education research offers a path forward. We release BenchMarker to bridge the fields and improve MCQA benchmark design."
}