Summary of Can Large Language Models Make the Grade? An Empirical Study Evaluating Llms Ability to Mark Short Answer Questions in K-12 Education, by Owen Henkel et al.
Can Large Language Models Make the Grade? An Empirical Study Evaluating LLMs Ability to Mark Short Answer Questions in K-12 Education
by Owen Henkel, Adam Boxer, Libby Hills, Bill Roberts
First submitted to arxiv on: 5 May 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary A novel dataset is used to evaluate the performance of Large Language Models (LLMs) in marking open text responses to short answer questions. The study explores how well different combinations of GPT versions and prompt engineering strategies perform at marking real student answers across various domain areas and grade levels using a Carousel quizzing platform dataset. Results show that GPT-4 with basic few-shot prompting performs well, with a Kappa score of 0.70 and human-level performance close to 0.75. This research builds on prior findings that GPT-4 can reliably score short answer reading comprehension questions at expert human rater levels. The study suggests that LLMs could be valuable tools for supporting low-stakes formative assessment tasks in K-12 education. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper shows how well computers can grade student answers to short questions. Researchers tested different types of computer models and ways to ask questions, using a big collection of real student answers from an online quiz platform. They found that one type of model, GPT-4, is very good at grading these answers, almost as good as human teachers! This could be important for schools because it might help computers do some teaching tasks, like giving feedback on student work. |
Keywords
» Artificial intelligence » Few shot » Gpt » Prompt » Prompting
