The Divided Team
A manager leads a 12-person product team. In a meeting, she asks for opinions on a risky launch. The two most senior members speak first and both favor launching immediately. As the discussion proceeds, several quieter members glance at the seniors before answering, and the group converges on 'launch now' with little dissent. Privately, afterward, three members tell the manager they had serious doubts but did not want to 'be the problem.' Separately, the manager has noticed that team members who work in the same open-plan area have started to share the same pessimistic view of the launch's odds, while a remote subgroup remains cautiously optimistic. A junior member who raised a concern last month was teased as 'the worrier' and has been quiet since.
1. Identify the group processes at work in the meeting and distinguish them from one another.
Read the worked answer
Groupthink (Janis): The drive for consensus suppresses dissent. Symptoms visible here: self-censorship (quiet members withholding doubts), pressure on dissenters (the 'worrier' teasing), illusion of unanimity (everyone says launch), and mindguards/authority deference (seniors speak first).
Conformity / normative social influence: Members publicly agree to avoid standing out, even while privately doubting - exactly what the three private disclosures show. This is compliance without internalization.
Informational social influence: Some may genuinely update toward the seniors' view because they assume the seniors know more. This differs from normative influence: one changes private belief, the other only public behavior.
Group polarization: Discussion among like-minded people shifts the group toward a more extreme position than members held individually - here, toward a confident 'launch now.'
Distinguishing them matters: conformity is about pressure to match others; polarization is about the discussion moving the average; groupthink is the structural failure that lets both go unchecked.
2. The open-plan and remote subgroups now hold different views. Explain this using group-level mechanisms, and say which mechanism is most likely responsible.
Read the worked answer
Most likely: group polarization within the co-located subgroup, amplified by shared information. People who interact frequently and hear mainly each other's arguments rehearse and strengthen the dominant view. The open-plan group's pessimism becomes more extreme through repeated discussion; the remote group, interacting less with the pessimists and more with each other, drifts the other way.
Contributing mechanisms:
- Informational influence within subgroups: each subgroup hears a biased sample of arguments, so its members update in the locally dominant direction.
- Social identity / in-group favoritism: proximity creates an 'us' (in-office) versus 'them' (remote) distinction, and people conform more to their own subgroup.
- Biased information sampling: co-located members share the same pessimistic anecdotes and treat them as consensus evidence.
Least likely: simple personality differences. The pattern tracks physical arrangement, not individual traits, which points to group-level processes.
3. Propose three concrete structural changes the manager could make to improve the quality of the decision, and explain the mechanism behind each.
Read the worked answer
1. Collect independent judgments before discussion (e.g., written anonymous pre-votes). Mechanism: prevents anchoring on the first/most senior speaker and blocks normative pressure, so private doubts surface. It also gives the manager a baseline to compare against the post-discussion shift, revealing polarization.
2. Assign a rotating devil's advocate or use a 'pre-mortem' ('assume the launch failed - why?'). Mechanism: legitimizes dissent so it no longer carries social cost (reduces the 'worrier' stigma) and forces the group to generate disconfirming arguments the consensus would otherwise suppress.
3. Have the leader withhold her own opinion until the end and explicitly invite minority views; ensure remote members contribute through the same structured channel. Mechanism: reduces authority-driven conformity and equalizes participation across subgroups, countering the in-office/remote information split.
Optional: mixed-subgroup discussion or a structured debate to break the biased sampling. The common thread is changing the structure of information flow rather than exhorting people to 'speak up,' which rarely overcomes normative pressure.