In the educational landscape, collaborative assignments play a crucial role in fostering student interactions and knowledge
exchange and are ever more common in recent years. However, the personal factors influencing team formation and the impact of
relationships dynamics on academic performance is complex and still unknown. With this in mind, we designed and conducted
an experiment across multiple schools in the Barcelona region to delve into the intricate relationship between student dynamics
and academic performance. Our research particularly focuses on the influence of friendship dynamics and individual characteristics
on team formation and group performance among students.

The first objective of our study is to investigate how student groups are formed and how students choose their peers based
on different characteristics such as ability, bullying, gender, or friendship. Our findings reveal significant homophily patterns in
gender, and ability is an important factor to be chosen into someone else’s group. A concerning result is that students experiencing
bullying are less likely to be preferred in the groups of their peers.

The second objective focuses on assessing how group composition affects student performance. We employ two distinct
metrics, mathematics and logic, to evaluate student performance under different grouping policies. By doing so, we gain insights
into the impact of group dynamics on academic outcomes. Additionally, our research explores how teams would be formed and
how student performance would be affected if teachers were responsible for creating the groups. This comparative analysis sheds
light on the potential benefits or drawbacks of various group formation methods within the educational setting.