Learning alone or in a group — An empirical case study of the collaborative learning patterns and their impact on student grades

Collaborative learning continues to thrive on the back of the rapid growth in cheap and powerful knowledge access technologies connecting and enabling students to carry out ever more learning, coursework and assessment tasks together. This work reports on the Collaborative Learning Environment (CLE) platform developed at EBTIC and tried along one semester for various groups of 122 students from the Engineering and Molecular Biology courses at Khalifa University. Firstly, we intend to uncover different patterns of student collaborations along the sequence of 3 courseworks and explore their impact on student learning outcome. Secondly, a generative mixture model is proposed to isolate the impact of the sheer collaboration from individual students qualities on the group performance, which assumes that the group coursework grade is generated as a combination of fixed students' grade expectations improved or damaged by the collaboration type that the students chose to follow. The results from this case study indicate that group learning with efficient collaboration patterns improves students' performance comparing to learning alone.