Machine-learning-based software to group heterogeneous students for online peer assessment activities

Since the academic year 2017/2018, a peer assessment activity was included in the online Genomics laboratory for the master’s degree course in Biological Sciences of the University of Camerino, with the aim of improving learning outcomes and soft skills in students, such as team building and critical thinking. Creating groups in university courses is not easy because of the large number of students, that leads teachers to realize groups totally randomly, a procedure that is not always effective. One of the factors that influences the success of collaborative learning is the creation of heterogeneous groups based on the students’ behaviors. Despite little improvements, the online genomics laboratory highlighted some gaps. Random groups didn’t ensure that each group was composed of heterogeneous students, and it leads some students to have a bad perception of the peer review activity, negatively affecting their engagement and motivation. This work proposes a new Machine Learning Approach and the realization of a specific software, able to create effective heterogeneous groups to be involved in the online peer assessment process, in order to improve learning outcomes and satisfaction in the students. The aim is to check the improvement of the peer assessment effectiveness using heterogeneous groups compared to random groups of students. Two editions of the online laboratory of Genomics were analysed, examining the students’ results and perceptions to verify the impact of the Machine Learning approach designed in this work.

This version of the paper has been accepted for publication, after peer review and is subject to Springer Nature’s AM terms of use (https://www.springernature.com/gp/open-research/policies/accepted-ma...), but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: http://dx.doi.org/10.1007/978-3-031-29800-4_2