Collaborative Learning Group Formation Based on Personality Traits: An Empirical Study in Initial Programming Courses

Abstract

Considering that the group formation is one of the key processes when developing activities in collaborative learning scenarios, the aim of this paper is to propose a technique based on an approach of genetic algorithms to achieve homogeneous groups, considering the students’ personality traits as grouping criteria. The main feature of this technique is that it allows the consideration of as many traits of the student as desired, converting the grouping problem in one of multi-objective optimization, given the combinatorial explosion that can occur depending on the number of students and of groups. For its validation, an experiment was designed with 132 first semesters engineering students, quantifying their personality traits through the “Big Five Inventory”, forming work groups and developing a collaborative activity in initial Programming courses. The experiment made it possible to compare the results obtained by the students applying the proposed approach to those obtained through other group formation strategies. It was demonstrated through the experiment that the homogeneous groups generated by the proposed technique produced better academic results compared to the formation techniques traditionally used by the teachers, when developing a collaborative activity.