A teacher wants to divide 30 students of her class in small groups for a task…

A group's usefulness for a task depends on matching its composition to the task's purpose. When the goal is to build understanding of diversity, the groups themselves must be heterogeneous — deliberately mixing students across backgrounds, abilities and social circles — because a homogeneous group, formed by comfort, seating order or exam rank, mainly reproduces a student's existing circle and gives no exposure to people unlike themselves.

Here the task is specifically about diversity, so the students in each group need to differ from one another. Assigning heterogeneous groups at random through a game mixes students who would not otherwise team up, giving every group members with different backgrounds, viewpoints and ways of thinking — while the random, game-based method keeps the process fair and avoids singling anyone out by ability or background.

Checking each alternative against this goal shows why it falls short:

• Letting students group by comfort mostly reproduces existing friend circles, which is the opposite of the mix a diversity task needs.

• Grouping benchwise, front to back, only reflects where students happen to sit — a seating arrangement, not a deliberate mix of backgrounds.

• Ranking students by their last exam marks clusters similar academic performers together, producing academic sameness rather than diversity.