Distilling ChatGPT for Explainable Automated Student Answer Assessment

TL;DR

This paper presents a method to distill ChatGPT into a smaller model for explainable automated student answer assessment, improving scoring accuracy and rationale quality.

Contribution

The paper introduces a novel framework that fine-tunes a smaller language model using ChatGPT-generated rationales for better explainability and accuracy in student answer assessment.

Findings

11% improvement in QWK score over ChatGPT

Generated rationales are comparable to ChatGPT's

Effective approach for explainable automated assessment

Abstract

Providing explainable and faithful feedback is crucial for automated student answer assessment. In this paper, we introduce a novel framework that explores using ChatGPT, a cutting-edge large language model, for the concurrent tasks of student answer scoring and rationale generation. We identify the appropriate instructions by prompting ChatGPT with different templates to collect the rationales, where inconsistent rationales are refined to align with marking standards. The refined ChatGPT outputs enable us to fine-tune a smaller language model that simultaneously assesses student answers and provides rationales. Extensive experiments on the benchmark dataset show that the proposed method improves the overall QWK score by 11% compared to ChatGPT. Furthermore, our thorough analysis and human evaluation demonstrate that the rationales generated by our proposed method are comparable to those of ChatGPT. Our approach provides a viable solution to achieve explainable automated assessment in education. Code available at https://github.com/lijiazheng99/aera.