A custom-designed chatbot in blended learning environments shows promise for student engagement but faces challenges in comprehensive coverage of academic topics and maintaining consistent student interaction.

Objective: To examine the effectiveness of a custom-designed chatbot in a blended learning environment, focusing on its adherence to instructional guidelines and impact on student engagement for academic major selection guidance.

Methods:

• Implementation in a first-year undergraduate business program
• Approximately 600 students across 31 sections
• Analysis of 64 valid chat sessions
• Quantitative content analysis (QCA) of chatbot interactions
• Custom GPT-based chatbot named "Career Guru" developed using ChatGPT's customizable features
• Students explored six academic majors through chatbot interactions

Key Findings:

• Uneven coverage of academic majors (FIN: 92.2%, MKT: 84.4%, DBM: only 31.3%)
• 32.8% of sessions showed chatbot pushing for more detailed answers
• 37.3% of conversations showed extended engagement after recommendations
• Some students disengaged when pushed for more detailed responses
• High success rate in providing recommendations in completed sessions
• Three conversations showed clear student enjoyment and genuine interest

Implications:

• Demonstrates potential for chatbots in academic guidance
• Shows need for simplified task design
• Highlights importance of face-to-face introductory sessions
• Supports integration of AI tools in blended learning environments
• Suggests value of personalized interactions in academic advising

Limitations:

• Small sample size (only 10% of possible sessions analyzed)
• Varying instructor experience with blended learning
• Student preferences not considered in initial design
• Lack of consistent implementation across sections

Future Directions:

• Focus on streamlining chatbot tasks
• Better integration of student preferences
• Design more cohesive learning experiences
• Improve connection between in-class and online interactions
• Explore implementation in various educational settings

Title and Authors: "Supporting Student Engagement and Academic Guidance through a Chatbot in Blended Learning" by Vasa Buraphadeja and Donyawan Chantokul

Published on: 2024

Published by: International Journal on E-Learning (2024) 23(3), 241-253

The study provides valuable insights into the practical implementation of AI chatbots in educational settings, particularly for academic guidance and student engagement in blended learning environments. While showing promise, it also highlights important considerations for future implementations, such as the need for simpler task design and better integration with face-to-face components.