TAing in the era of GenAI: Establishing guidelines to support use of Generative AI for feedback generation by teaching assistants

A robot helps a mouse grade student assignments by providing feedback using Generative AI

Grant Recipients

Kenneth McKay, Ada Hurst, Jackie Heitzner, Management Science and Engineering

(Project timeline: September 2024 - August 2025)

Description

This project aims to enhance the effectiveness of Teaching Assistants (TAs) by utilizing Generative AI (GenAI) to generate feedback on student submissions. The project acknowledges the potential of GenAI to enhance feedback depth and insightfulness, providing TAs with strategies for feedback. This grant will fund Phase 1 of the project, which addresses key questions about identifying the best practices, tools, and resources for GenAI in assessment and determining necessary adaptations for TA use. This work comprises three key activities. First, current uses of GenAI in student assessment are reviewed in the relevant academic and grey literature. Second, the use of GenAI for generating feedback is tested in a non-STEM course, using mock student answers and TA assessments. Third, building on insight from the previous activities, guidelines are developed for TAs to use GenAI effectively. Phase 2, beyond the current proposal, will integrate these guidelines into TA training. 

Project Objectives

At a high level this project concerns supporting TAs in effectively utilizing advances in GenAI in their duties, particularly in developing individualized formative and summative feedback on student submissions. The project comprises two main phases: 

The objective of Phase 1 is to collect, explore and develop best practices to support TA’s use of GenAI in crafting feedback. We anticipate this will comprise three main outcomes: 

  • O1. Complete a review of current uses of GenAI in supporting student assessment.  
  • O2. Explore and test the use of GenAI in grading and generating feedback on assessments. 
  • O3. Develop a set of guidelines to support TAs in using GenAI for student feedback generation. 

Accordingly, these activities will investigate the following questions: 

  • Q1. What are the best practices, tools, and resources for using GenAI in student assessment?  
  • Q2. To what extent can GenAI be used to support feedback generation? 
  • Q3. What adaptations are required to apply these methods to support TAs in generating student feedback? 

Phase 2 of the project will see the outcome of Phase 1 (the produced guidelines) being progressively incorporated in TA training.