AI-Enabled grading with near-domain data for scaling feedback with human-level accuracy

TL;DR

This paper introduces a practical AI-based framework for grading short-answer questions using near-domain data, outperforming existing models and large language models, with no need for pre-written rubrics, thus enabling scalable and accurate feedback in education.

Contribution

It formalizes the problem of automated short answer grading using near-domain data and presents a novel framework that surpasses state-of-the-art models and large language models in accuracy.

Findings

Our method outperforms state-of-the-art ML models by 10-20%.

Large language models like GPT 4 perform significantly worse without fine-tuning.

Near-domain data enhances grading accuracy and provides valuable learning insights.

Abstract

Constructed-response questions are crucial to encourage generative processing and test a learner's understanding of core concepts. However, the limited availability of instructor time, large class sizes, and other resource constraints pose significant challenges in providing timely and detailed evaluation, which is crucial for a holistic educational experience. In addition, providing timely and frequent assessments is challenging since manual grading is labor intensive, and automated grading is complex to generalize to every possible response scenario. This paper proposes a novel and practical approach to grade short-answer constructed-response questions. We discuss why this problem is challenging, define the nature of questions on which our method works, and finally propose a framework that instructors can use to evaluate their students' open-responses, utilizing near-domain data like data from similar questions administered in previous years. The proposed method outperforms the state of the art machine learning models as well as non-fine-tuned large language models like GPT 3.5, GPT 4, and GPT 4o by a considerable margin of over 10-20% in some cases, even after providing the LLMs with reference/model answers. Our framework does not require pre-written grading rubrics and is designed explicitly with practical classroom settings in mind. Our results also reveal exciting insights about learning from near-domain data, including what we term as accuracy and data advantages using human-labeled data, and we believe this is the first work to formalize the problem of automated short answer grading based on the near-domain data.