Group Size and Peer Learning: Peer Discussions in Different Group Size Influence Learning in a Biology Exercise Performed on a Tablet With Stylus

Résumé En

Determining the optimal discussion group size to improve performance and learning has created an intense debate in psychology and provided mixed findings in laboratory and field settings. In a quasi-experimental study in the education field, we examined the effect of discussion group size on individual learning in a biology exercise performed on a tablet with stylus. The sample involved 102 secondary school students divided into four classes, each corresponding to one of the four experimental conditions (alone without peer discussion vs discussion in dyads vs triads vs four-member groups). They were asked to draw individually a functional schema of the human respiratory system, once before and once after discussing with peers (or reflecting alone). Both drawings were evaluated by four independent coders, and the learning gain for each student was computed from these evaluations. Results revealed that learning gain was greater for students discussing in four-member groups than for those in the other conditions. Additional analyses suggested that this effect was moderated by the students' gender, with the learning gain being greater after discussion in four-member groups among females only. These findings suggest that group size of four individuals might be the optimal configuration to improve peer learning.