Concurrent Session #3: Using Generative AI to Efficiently Create and Modify Assessment Problems

Alternative Title

Using Generative Artificial Intelligence (GenAI) to Efficiently Create and Modify Assessment Problems

Contributor

University of Central Florida. Faculty Center for Teaching and Learning; University of Central Florida. Division of Digital Learning; Teaching and Learning with AI Conference (2023 : Orlando, Fla.)

Location

Cape Florida B

Start Date

24-9-2023 2:00 PM

End Date

24-9-2023 2:15 PM

Publisher

University of Central Florida Libraries

Keywords:

Assessment generation; STEM education; AI in education; Isomorphic problems; Misinformation detection

Subjects

Artificial intelligence--Educational applications; Artificial intelligence--Computer-assisted instruction; Physics--Computer-assisted instruction; Problem solving--Computer-assisted instruction; Generative Exam System (Computer system)

Description

Creating high quality assessment problems is a highly time-consuming task for instructors, especially in STEM disciplines such as math and physics. With proper prompt engineering, generative AI such as ChatGPT can be used to either generate new problems or create isomorphic variations of existing problems efficiently. I will talk about how large language models such as GPT-3.5 can be used to quickly generate large numbers of isomorphic problems with figures and solutions in physics, which are used to transform classroom exams. I will also share experiences in creating new types of STEM assessments that measure students’ ability to identify and argue against common misinformation on the internet.

Language

eng

Type

Presentation

Rights Statement

All Rights Reserved

Audience

Faculty

This document is currently not available here.

Share

COinS
 
Sep 24th, 2:00 PM Sep 24th, 2:15 PM

Concurrent Session #3: Using Generative AI to Efficiently Create and Modify Assessment Problems

Cape Florida B

Creating high quality assessment problems is a highly time-consuming task for instructors, especially in STEM disciplines such as math and physics. With proper prompt engineering, generative AI such as ChatGPT can be used to either generate new problems or create isomorphic variations of existing problems efficiently. I will talk about how large language models such as GPT-3.5 can be used to quickly generate large numbers of isomorphic problems with figures and solutions in physics, which are used to transform classroom exams. I will also share experiences in creating new types of STEM assessments that measure students’ ability to identify and argue against common misinformation on the internet.