pAIrfeedL

pAIrfeed-L

Implementation of Large Language Models
in Dialogue-Based Peer Feedback and Investigation of Its Effects

This project is supported by TUBITAK within the scope of 3501 - Career Development Program. Project No: 323K213

Summary

Summary of the Project

Importance

In order to catch the age of artificial intelligence in education in our country, the proposed project will use artificial intelligence to support feedback, one of the most important components of learning, and will investigate its effects on learning.
The scientific evidence obtained from experimental studies during the project will contribute to the establishment of a competent and competitive education system in the field of artificial intelligence by guiding policies of higher education.

Research Questions

What training principles and machine learning techniques are required to train artificial intelligence as a peer giving feedback?
Can artificial intelligence peers create a greater impact on feedback quality and learning compared to human student peers?
What are the effects of artificial intelligence peers who can give feedback adaptively according to the knowledge levels of students on feedback quality and learning compared to non-adaptive artificial intelligence peers?

Originality

The original value of this project is related to the discovery of new information and approaches about the feedback processes enabled by artificial intelligence peers and their effects on learning as well as about the artificial intelligence-student interaction.
In addition, the effects of adaptive artificial intelligence models that can provide feedback in accordance with the individual differences and knowledge levels of students will be investigated for the first time with this study.

Goals and Objectives

Effective Use of Artificial Intelligence in Education

The proposed project aims to advance peer feedback applications by effectively integrating artificial intelligence into university education.

The project aims to offer recommendations based on scientific findings for policies aimed at the correct and effective integration of artificial intelligence into education.

Impact

The role of LLM in Education

This project will produce conference papers, presentations, and articles related to pilot and experimental studies to be carried out in various work packages. The peer feedback system will be opened for use by other universities. Various trainings and workshops related to artificial intelligence will be organized at universities.
This project will personalize students’ learning processes through the artificial intelligence-supported feedback system and contribute to increasing quality and accessibility in education. The project results will be shared with important institutions in Turkey (for example, YÖK and MEB), guiding the policies of implementing artificial intelligence in education nationwide.

Method

Method of the Project

This project will address the research questions with both quantitative and qualitative methods using mixed-methods triangulation. Experimental studies and descriptive-descriptive approaches will be used to measure the effects of artificial intelligence support. The study will be conducted with the approval of the METU Human Research Ethics Committee.