# LLMs for Qualitative Data Analysis Fail on Security-specificComments in Human Experiments

arXiv: https://arxiv.org/abs/2604.10834 · HTML: https://mlsanity.com/paper/2604.10834/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 12 of its 63 references that are in the map.
- First posted on arXiv 12 Apr 2026.

## Builds on (most cited first, 12 of 12)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Visualizing and Understanding Convolutional Networks](https://mlsanity.com/paper/1311.2901/) (2013)
- [Supporting Qualitative Analysis with Large Language Models: Combining Codebook with GPT-3 for Deductive Coding](https://mlsanity.com/paper/2304.10548/) (2023)
- [LLM-in-the-loop: Leveraging Large Language Model for Thematic Analysis](https://mlsanity.com/paper/2310.15100/) (2023)
- [Scalable Qualitative Coding with LLMs: Chain-of-Thought Reasoning Matches Human Performance in Some Hermeneutic Tasks](https://mlsanity.com/paper/2401.15170/) (2024)
- [Large Language Models in Qualitative Research: Uses, Tensions, and Intentions](https://mlsanity.com/paper/2410.07362/) (2024)
- [Predicting Usefulness of Code Review Comments using Textual Features and Developer Experience](https://mlsanity.com/paper/1807.04485/) (2018)
- [Sentiment Polarity Detection for Software Development](https://mlsanity.com/paper/1709.02984/) (2017)
- [EmoTxt: A Toolkit for Emotion Recognition from Text](https://mlsanity.com/paper/1708.03892/) (2017)
- [It's Difficult to be Neutral -- Human and LLM-based Sentiment Annotation of Patient Comments](https://mlsanity.com/paper/2404.18832/) (2024)
- [On the acceptance by code reviewers of candidate security patches suggested by Automated Program Repair tools](https://mlsanity.com/paper/2209.07211/) (2022)

## Abstract (arXiv)

[Background:] Thematic analysis of free-text justifications in human experiments provides significant qualitative insights. Yet, it is costly because reliable annotations require multiple domain experts. Large language models (LLMs) seem ideal candidates to replace human annotators. [Problem:] Coding security-specific aspects (code identifiers mentioned, lines-of-code mentioned, security keywords mentioned) may require deeper contextual understanding than sentiment classification. [Objective:] Explore whether LLMs can act as automated annotators for technical security comments by human subjects. [Method:] We prompt four top-performing LLMs on LiveBench to detect nine security-relevant codes in free-text comments by human subjects analyzing vulnerable code snippets. Outputs are compared to human annotators using Cohen's Kappa (chance-corrected accuracy). We test different prompts mimicking annotation best practices, including emerging codes, detailed codebooks with examples, and conflicting examples. [Negative Results:] We observed marked improvements only when using detailed code descriptions; however, these improvements are not uniform across codes and are insufficient to reliably replace a human annotator. [Limitations:] Additional studies with more LLMs and annotation tasks are needed.