4328-GRP ProgCT: Cultivating Computational Thinking with Pedagogy-Grounded AI Agents

Presenter Information

Location

https://campus.kennesaw.edu/colleges-departments/ccse/events/computing-showcase/fa26-cday-program.php

Document Type

Event

Start Date

20-11-2026 2:00 PM

Description

Computational Thinking (CT) develops critical thinking and problem-solving skills in learners systematically. Modern AI tutors can help learners solve problems effectively; however, they generate direct answers and mostly fail to integrate the pedagogical principles focused on CT components to improve students’ problem-solving skills. This failure raises concern in the development of CT skills for early learners. Our project, ProgCT is a pedagogy-grounded multi-agent tutor that helps learners develop computational problem-solving abilities through CT practices. As learners work through each problem, the agents (i) assess the CT skills learners demonstrate, (ii) analyze learners’ cognitive engagement, and (iii) guide learners through the relevant CT steps with Socratic hints without revealing the solution. In a pretest post-test study on Python programming concepts involving seven learners (n=7), the mean score of the Algorithmic Thinking Test for Adults (ATTA) rose from 64.8% to 77.6%, and self-assessed confidence on the Hi-ACT test increased in all six CT components – abstraction, decomposition, pattern recognition, algorithm design, evaluation and generalization. The system’s built-in progress evaluation module also recorded increases in participants’ CT ownership and cognitive engagement on later exercises compared with earlier exercises of the same difficulty.

This document is currently not available here.

Share

COinS
 
Nov 20th, 2:00 PM

4328-GRP ProgCT: Cultivating Computational Thinking with Pedagogy-Grounded AI Agents

https://campus.kennesaw.edu/colleges-departments/ccse/events/computing-showcase/fa26-cday-program.php

Computational Thinking (CT) develops critical thinking and problem-solving skills in learners systematically. Modern AI tutors can help learners solve problems effectively; however, they generate direct answers and mostly fail to integrate the pedagogical principles focused on CT components to improve students’ problem-solving skills. This failure raises concern in the development of CT skills for early learners. Our project, ProgCT is a pedagogy-grounded multi-agent tutor that helps learners develop computational problem-solving abilities through CT practices. As learners work through each problem, the agents (i) assess the CT skills learners demonstrate, (ii) analyze learners’ cognitive engagement, and (iii) guide learners through the relevant CT steps with Socratic hints without revealing the solution. In a pretest post-test study on Python programming concepts involving seven learners (n=7), the mean score of the Algorithmic Thinking Test for Adults (ATTA) rose from 64.8% to 77.6%, and self-assessed confidence on the Hi-ACT test increased in all six CT components – abstraction, decomposition, pattern recognition, algorithm design, evaluation and generalization. The system’s built-in progress evaluation module also recorded increases in participants’ CT ownership and cognitive engagement on later exercises compared with earlier exercises of the same difficulty.