Skip to content

ChatGPT-based automated HDR assessment system (problem generation, scoring, and feedback pipeline)

Design · All designs
Evidence · no claims cited

Description

An LLM-based automated environment built with ChatGPT (GPT-4o) that integrates three prompts—problem generation, scoring, and feedback—into one system handling the full HDR assessment cycle. The scoring and feedback prompts are integrated into one prompt, with "Explanation" and "Good Points" sections outputting the score rationale and feedback. The system generates problems aligned with educational intentions and quickly assesses descriptive tasks focused on knowledge misuse, with reproducibility, immediacy, and low cost.

Design Implications

Context

Requirements

  • A pretest collecting actual descriptive answers from learners to serve as the basis for scoring and feedback prompts
  • Prompts designed for consistency, educational validity, and explainability

Constraints

  • Some problem and answer data were collected in a previous study (Takahashi et al., 2025)
  • Validated only with GPT-4o; behavior may differ across LLMs

Target Learners

  • Working adults in business education (MBA-style programs)

Learning Goals

  • Detection and explanation of knowledge misuse in practical cases

Claims

Examples

-

Key Sources

  • Satoshi Takahashi, Atsushi Yoshikawa, Megumi Kose, Kenichi Suzuki, Chieko Inoue, Yumi Watanabe, and Mari Sawada. (2026). Automating Constructive Assessment with Large Language Models: Toward Scalable and Repeated Evaluation of Practical Competence. arXiv. https://arxiv.org/abs/2609.25790
Page metadata
Field Value
type design
id chatgpt-automated-hdr-assessment-system
title ChatGPT-based automated HDR assessment system (problem generation, scoring, and feedback pipeline)
description An LLM-based automated environment built with ChatGPT (GPT-4o) that integrates three prompts—problem generation, scoring, and feedback—into one system handling the full HDR assessment cycle.
status review
generated by process:settle-candidates on 2026-10-10
trust unverified — no one has confirmed this page's content
sources 1 — satoshi-takahashi-2026

Frontmatter is the page's machine-readable record — see the Schema & Guide.