Agreement, Calibration, and Failure of Three Large Language Models as High-Stakes Multimodal OSPE Graders: A Comparative Psychometric Analysis

Read full text Summarise with VIRAT
Authors: Sadia Qazi
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This supplementary file presents three representative OSPE stations illustrating large language model grading behavior. Each example includes the standardized grading prompt, the OSPE image, and the raw outputs generated independently by Claude 3.5 Haiku, Gemini 2.5 Flash, and ChatGPT-4o. Examples were selected to demonstrate typical grading logic and observed divergence patterns. All student identifiers and institutional metadata have been removed.

Fields of Study & Subjects

Artificial Intelligence in Healthcare and Education Innovations in Medical Education Artificial Intelligence in Education

Publication details

Journal / source
Figshare
Publication date
1 January 2026
Language
EN