Grouping students appropriately to increase learning achievement is important in learning and teaching. Traditional grouping methods include both homogeneous and heterogeneous grouping; heterogeneous grouping has been claimed to improve students' learning achievement and learning process in both cooperative and collaborative learning. Recently, machine–learning-based grouping approaches have been proposed to produce better heterogeneous groups. One main drawback of these machine-learning-based methods is that they are highly affected by parameter settings; setting the appropriate parameters is difficult for general users. Consequently, the most adopted heterogeneous grouping methods currently are s-shape placement, random assignment, and self-grouping, as the three methods do not require additional parameter settings. Herein, a new heterogeneous grouping algorithm named MASA (magic square-based heterogeneous grouping algorithm) is proposed. As in the s-shape placement method, the only parameter required in MASA is the number of groups. Experimental analysis on 92 datasets indicated that MASA was superior to the s-shape placement, random assignment, and self-grouping methods for generating better heterogeneous groups. Additionally, MASA is an adaptive method that can generate several grouping results simultaneously, and users can select the preferred solution.
All Science Journal Classification (ASJC) codes
- Control and Systems Engineering
- Theoretical Computer Science
- Computer Science Applications
- Information Systems and Management
- Artificial Intelligence