# Comparison of Scoring Rationales Between Large Language Models and Human Raters

arXiv: https://arxiv.org/abs/2509.23412 · HTML: https://mlsanity.com/paper/2509.23412/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 24 references that are in the map.
- First posted on arXiv 27 Sep 2025.

## Builds on (most cited first, 8 of 8)

- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Fine-tuning ChatGPT for Automatic Scoring](https://mlsanity.com/paper/2310.10072/) (2023)
- [Human-AI Collaborative Essay Scoring: A Dual-Process Framework with LLMs](https://mlsanity.com/paper/2401.06431/) (2024)
- [Language models and Automated Essay Scoring](https://mlsanity.com/paper/1909.09482/) (2019)
- [Automated Essay Scoring Using Transformer Models](https://mlsanity.com/paper/2110.06874/) (2021)
- [Beyond Agreement: Diagnosing the Rationale Alignment of Automated Essay Scoring Methods based on Linguistically-informed Counterfactuals](https://mlsanity.com/paper/2405.19433/) (2024)
- [Calibrating LLMs with Preference Optimization on Thought Trees for Generating Rationale in Science Question Scoring](https://mlsanity.com/paper/2406.19949/) (2024)
- [Investigating Automatic Scoring and Feedback using Large Language Models](https://mlsanity.com/paper/2405.00602/) (2024)

## Built on by (most cited first, 1 of 1)

- [Exploration of Summarization by Generative Language Models for Automated Scoring of Long Essays](https://mlsanity.com/paper/2510.22830/) (2025)

## Abstract (arXiv)

Advances in automated scoring are closely aligned with advances in machine-learning and natural-language-processing techniques. With recent progress in large language models (LLMs), the use of ChatGPT, Gemini, Claude, and other generative-AI chatbots for automated scoring has been explored. Given their strong reasoning capabilities, LLMs can also produce rationales to support the scores they assign. Thus, evaluating the rationales provided by both human and LLM raters can help improve the understanding of the reasoning that each type of rater applies when assigning a score. This study investigates the rationales of human and LLM raters to identify potential causes of scoring inconsistency. Using essays from a large-scale test, the scoring accuracy of GPT-4o, Gemini, and other LLMs is examined based on quadratic weighted kappa and normalized mutual information. Cosine similarity is used to evaluate the similarity of the rationales provided. In addition, clustering patterns in rationales are explored using principal component analysis based on the embeddings of the rationales. The findings of this study provide insights into the accuracy and ``thinking'' of LLMs in automated scoring, helping to improve the understanding of the rationales behind both human scoring and LLM-based automated scoring.