Explain group development and prevent faulty decisions

Tuckman and Jensen describe five possible stages

Forming: people meet, learn the task and look for direction. Storming: differences, goals or ways of working become visible. Norming: members develop shared expectations and ways to coordinate. Performing: attention can focus on the task with more established collaboration. Adjourning: a temporary group finishes, hands over work or separates.

The stages are a useful description, not a timetable every group must follow. A team may skip, revisit or experience several patterns at once. A new member, changed task or conflict can disrupt established routines. Diagnose the group from evidence about its task and relationships instead of assuming its stage from its age.

Belbin groups nine ways people can contribute

Social roles are Resource Investigator (explores outside ideas and contacts), Teamworker (supports cooperation) and Co-ordinator (clarifies goals and draws on others’ strengths). Thinking roles are Plant (offers original ideas), Monitor Evaluator (analyses options) and Specialist (contributes focused expertise). Action roles are Shaper (drives progress), Implementer (turns plans into practical work) and Completer Finisher (checks detail and completion).

These are clusters of behaviour that can help a team progress, not job titles, diagnoses or fixed personality boxes. A person may contribute in several ways, and a role that is useful at one stage may be less useful at another. A balanced team considers the task and missing contributions rather than ranking people or treating a questionnaire as destiny.

Groupthink is pressure for agreement that weakens appraisal

Janis used groupthink to describe a decision process in a cohesive, insulated group where pressure to preserve agreement can reduce critical evaluation. Possible symptoms include an illusion of invulnerability, collective rationalisation, belief in the group’s inherent morality, negative stereotypes of outsiders, pressure on dissenters, self-censorship, an illusion of unanimity and ‘mindguards’ who filter inconvenient information.

An often-taught example is the US government’s Bay of Pigs decision, analysed retrospectively by Janis. It illustrates questions to ask about discussion and dissent; it does not prove that groupthink was the single cause. A poor outcome alone is not evidence of groupthink, and a cohesive team does not automatically make poor decisions. Look for the decision process and competing explanations.

Safeguards include a leader who withholds an initial preference, explicit invitations to disagree, independent subgroups, an assigned critical reviewer, outside expertise, checking disconfirming evidence and a second review before commitment. These steps reduce pressure but cannot guarantee a correct answer; the group still needs reliable information and clear criteria.

Forsyth groups decision errors by how information is mishandled

Sins of commission misuse information: belief perseverance keeps relying on evidence already shown to be inaccurate; sunk-cost bias preserves a failing plan because time or money has already been spent; extra-evidentiary bias uses information the group was told to exclude; hindsight bias overestimates how well an outcome could have been predicted beforehand.

Sins of omission leave out useful information: base-rate bias neglects relevant general-frequency evidence; the fundamental attribution error overemphasises a person’s disposition and underestimates situational causes; confirmation bias fails to seek evidence that could disconfirm the current conclusion.

Sins of imprecision use shortcuts that oversimplify: availability relies on examples that are easy to recall; conjunction bias treats a detailed combination of events as more likely than one of its parts; representativeness relies too heavily on a seemingly typical match. A label should name the information error and show how it affects this decision.

A real decision can contain more than one error. For example, after one vivid equipment failure a team might overestimate how common it is (availability), ignore base-rate records (omission) and keep an already funded but ineffective fix (sunk cost). Ask what information was used, missed or oversimplified before classifying the error.