AI-Based Adaptive Learning Model Reduces Cognitive Load in Vocational Students

Natalia Sihombing S.Pd Kamis, Oktober 08, 2026
Illustration by AI
FORMOSA NEWS

Novice students in vocational mechatronics education frequently experience high cognitive load during programming code debugging, which often triggers mental frustration and passive reliance on generative artificial intelligence. This study was developed and evaluated by Mustamin and Sirwan from the Department of Electronics Engineering Education at Universitas Negeri Makassar. Published in June 2026 in the International Journal of Educational Technology Research (IJETR), this research demonstrates how designing pedagogically integrated AI scaffolding can suppress extraneous cognitive load and foster autonomous problem-solving skills.

The background of this issue stems from the era of Industry 4.0 and Society 5.0, where mastering computer programming stands as a foundational yet complex competency for mechatronics students. When code errors occur, default compiler messages filled with technical jargon overburden working memory capacity, while unguided generative AI usage risks fostering passive copy-paste habits without grasping underlying computational principles.

The methodology employed a Research and Development (R&D) approach utilizing the ADDIE (Analysis, Design, Development, Implementation, Evaluation) instructional framework. This adaptive learning model was implemented across 8 practical laboratory sessions involving 21 active first-year undergraduate students enrolled in the Python Programming for Mechatronics course. Data were collected via post-activity quantitative questionnaires using a 5-point Likert scale, qualitative reflection sheets, and interaction matrix logs.

Based on data analysis, the primary findings of the study identify several key points:

  • Primary Sources of Cognitive Load: Logic errors accounted for 66.7% and indentation issues accounted for 42.9% as the major hurdles for novices.
  • Suppression of Passive Habits: Adaptive prompting guidance effectively mitigated passive copy-paste habits, yielding a low mean score of 2.33.
  • Boosted Independence: Independent debugging self-efficacy increased significantly with a mean score of 3.62 for autonomous error identification, where 76.2% of students performed independent analysis prior to consulting AI.
  • Task Completion Rates: A total of 85.7% of students successfully completed daily tasks and achieved autonomous code modification.
  • Psychological Affective Responses: Exactly 76.2% of students experienced cognitive relief and conceptual clarity, while 23.8% exhibited heightened intrinsic motivation.

The implications of this research prove that integrating AI scaffolding pedagogically within the Zone of Proximal Development (ZPD) successfully achieved Technology Readiness Level (TRL) 4. As emphasized by Mustamin and Sirwan from Universitas Negeri Makassar, this approach not only alleviates repetitive mental strain for instructors but also equips vocational students with adaptive and autonomous problem-solving skills in modern technology education.

Brief Author Profiles:

  • Mustamin, S.Pd., M.T. – Researcher and academic at the Department of Electronics Engineering Education, Universitas Negeri Makassar, focusing on mechatronics education, intelligent systems, and vocational learning technology.
  • Sirwan, S.Pd., M.Pd. – Researcher and academic at the Department of Electronics Engineering Education, Universitas Negeri Makassar, active in electronic engineering and hardware-software system integration.

Research Source:

Natalia Sihombing S.Pd

Diposting oleh Natalia Sihombing S.Pd

Posting Komentar

0 Komentar