A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
Group discussion tends to shift decisions toward more extreme positions than initial individual positions (group polarization): groups become more risk-taking or risk-averse depending on the initial lean. Social comparison (individuals competing to be most aligned with the group) and persuasive arguments (exposure to novel pro-attitudinal arguments) both drive polarization.
Simulate group decision-making where pre-discussion individual positions are measured, then measure post-discussion positions to quantify polarization; test whether polarization persists when social comparison motives are removed.
When researchers first studied group decision-making in the 1960s, they expected groups to moderate individual extremism — the conventional wisdom was that groups average out individual differences and produce cautious, centrist decisions. The empirical findings upended this assumption. James Stoner found in 1961 that groups discussing risky dilemmas made *riskier* decisions than the average of their individual members' pre-discussion positions. This became known as the risky shift. Further research revealed the pattern was more general: groups do not always shift toward risk — they shift toward the *direction already favored by their members*, and they shift further than any individual member started. This is group polarization: group discussion amplifies the dominant initial tendency of the group.
The phenomenon has two complementary mechanisms, and understanding both requires your prerequisite in social comparison theory. The first is social comparison: group members arrive with their own positions and observe where others stand. If you consider yourself moderately pro-environment and you discover that most group members are more strongly pro-environment than you, two things happen. First, you update your assessment of what the socially desirable position is. Second — and this is the key — you feel motivated to affirm your environmental values by moving further in the pro-environment direction, to distinguish yourself as genuinely committed rather than merely typical. The group's apparent consensus becomes a benchmark, and individuals compete to align with (or exceed) it. Social comparison does not just reveal the group's position; it triggers a race to endorse it more strongly.
The second mechanism is persuasive arguments: group discussion generates arguments, and in a group that already leans in a direction, the *preponderance of novel arguments will support that direction*. Before discussion, each person has already considered the most obvious arguments. Discussion surfaces arguments they had not personally generated — and because the group leans one way, statistically more new arguments favor that direction than oppose it. Hearing novel supporting arguments moves members further than they would have moved on private reflection alone. Both mechanisms predict the same outcome: a shift in the direction of the pre-existing lean, more extreme than the average starting position. The two mechanisms are independent and additive — each operates even when the other is controlled, and they reinforce each other in natural group discussion.
The practical stakes are significant. Group polarization helps explain why deliberative bodies sometimes reach more extreme outcomes than polling their members individually would suggest; why online communities that self-select around shared views become more radical over time; and why groupthink (your builds-toward topic) is not just about conformity pressure but about the inherent polarizing tendency of like-minded discussion. The crucial predictor of polarization direction is the *initial lean of the group*, not any feature of the discussion format itself. A risk-averse group discussing the same dilemma will polarize toward greater caution. This means polarization is not a bias toward one end of any scale — it is a bias that amplifies wherever you already are.