Arief Prasetyo, Aryo Pinandito, Tsukasa Hirashima
This study compares three concept mapping approaches in a basic computer networking course: concept map reading (CM-R), construction (CM-C), and recomposition (CMRec). Using a Latin square design with 40 Information Technology sophomores, we examined how these methods affect learning outcomes assessed through lower-order thinking questions (factual recall and basic conceptual understanding). This focus on foundational knowledge is essential for novice networking students before progressing to higher-order skills like network troubleshooting and design. Participants engaged in three networking topics across multiple sessions, experiencing all three approaches in counterbalanced sequences. We measured performance using pretests, posttests, and delayed tests, while cognitive load was evaluated through validated questionnaires. Results reveal distinct optimal applications: CM-R excelled in immediate learning gains with large effect sizes versus CM-Rec (p <0.001, r=0.620) and medium effects versus CM-C (p=0.038, r =0.327). However, CM-Rec demonstrated superior retention, achieving significantly higher retention ratios than both CM-R and CM-C, indicating effective knowledge consolidation. Cognitive load analysis shows CM-R minimizes extraneous load for efficient acquisition, while CM-Rec's high germane load supports deeper processing beneficial for retention. These findings suggest educators should strategically match methods to learning objectives: CM-R for efficient foundational knowledge acquisition, CM-Rec for sustained retention, and CM-C for deep constructive engagement when adequate support is available. © 2026 IEEE.
Hiroshima University, Graduate School of Advanced Science and Engineering, Hiroshima, Japan; Brawijaya University, Faculty of Computer Science, Information Systems Department, Malang, Indonesia
