Executive decision-making, bias and human factors
A premium course on how good leaders make bad decisions, and how to prevent it: cognitive biases and groupthink, normalisation of deviance, automation and human error, organisational silence, and practical tools such as the pre-mortem and decision reviews. Built on official investigations of Challenger, Columbia, Air France 447 and Mid Staffordshire, and on HSE’s HSG48. Not an accredited qualification.
What you will learn
- Recognise the cognitive biases and group dynamics that most often distort executive decisions.
- Explain normalisation of deviance using the Challenger and Columbia investigations.
- Apply human-factors thinking (HSG48) to errors, violations and automation, drawing on Air France 447.
- Identify organisational silence and target pressure, using the Mid Staffordshire inquiry.
- Use practical tools — pre-mortem, red team, decision log and review — to improve high-stakes decisions.
Modules
- How good leaders make bad decisions
- Normalisation of deviance: Challenger and Columbia
- Human error, violations and automation
- Organisational silence and target pressure
- Tools for better decisions
Course facts
- Estimated study time
- about 1.5 hours
- Final assessment
- 10 questions · Pass mark 80%
- Available in
- English, Français, العربية
- Price
- Free. Optional verifiable certificate: $12
- Last updated
- Published by
- AMAADOR ACADEMY
Who this course is for
Directors, executives and senior managers who make or approve high-stakes decisions, and those who advise them.
Prerequisites
No formal prerequisite.
Course content
1. How good leaders make bad decisions
Experienced leaders rely on judgement built from experience. That is usually a strength, but it brings predictable traps. Among the most relevant for executives: confirmation bias (seeking evidence that supports what we already believe), anchoring (over-weighting the first number or explanation), optimism bias and the planning fallacy (underestimating cost, time and risk), sunk-cost thinking (continuing because of what has already been spent), and overconfidence.
Groups can amplify these. Groupthink describes cohesive teams converging on a view without real challenge, especially under a strong leader, time pressure or external threat. Signs include an illusion of unanimity, self-censorship, dismissal of warnings and pressure on dissenters.
The answer is not to “try harder to be objective” — biases operate below awareness. It is to design decision processes that bring out alternative views and disconfirming evidence before commitment.
Key point: Speak last if you are the most senior person in the room.
2. Normalisation of deviance: Challenger and Columbia
Normalisation of deviance, a term coined by sociologist Diane Vaughan in her study of Challenger, describes how an organisation gradually comes to accept a departure from a standard because nothing bad has happened yet. Each success with the anomaly makes the next acceptance easier, until the margin is gone.
The space shuttle programme suffered this twice. Leaders asked the right question — “is it safe?” — but the burden of proof had silently shifted: engineers had to prove it was unsafe rather than managers proving it was safe.
Executives can spot drift by asking: which anomalies have we stopped reporting because they are “normal”? Which waivers or deviations have become permanent? Where do we rely on past success as evidence of safety?
Key point: “It has always been fine” is a description of the past, not evidence about the future.
Case studies: Challenger (1986) and Columbia (2003). Challenger broke apart shortly after launch on 28 January 1986, killing all seven crew. The Rogers Commission (June 1986) traced the loss to the failure of an O-ring seal in a solid rocket booster joint, and found the decision to launch was flawed: those who made it were unaware of the recent history of problems with the O-rings and of the contractor’s initial written recommendation against launching at low temperatures. Columbia disintegrated on re-entry on 1 February 2003, killing all seven crew, after foam struck the wing at launch. The Columbia Accident Investigation Board (August 2003) concluded that NASA’s organisational culture had as much to do with the accident as the foam, noting that foam strikes had come to be treated as an accepted, maintenance-level problem.
3. Human error, violations and automation
HSE’s guidance HSG48, Reducing error and influencing behaviour, distinguishes errors (slips, lapses and mistakes, which are unintentional) from violations (deliberate departures from rules, often well-intentioned or driven by pressure, poor procedures or impossible targets). Each needs a different response: better design, training and procedures for errors; understanding and removing the reasons for violations.
The guidance stresses that human failures are usually symptoms of organisational conditions — workload, equipment design, procedures, supervision and culture — rather than root causes. Blaming the individual closest to the event rarely prevents recurrence.
Automation adds a specific risk: when systems handle normal operations, people may lose practice and situational awareness, then face an unexpected handover in a crisis. Leaders deploying automation and AI should ask what happens when it fails, and whether people are trained for that moment.
Key point: When someone breaks a rule, first ask why the rule was hard to follow.
Case study: Air France 447 (2009). On 1 June 2009 an Airbus A330 operating flight AF 447 from Rio de Janeiro to Paris crashed into the Atlantic, killing all 228 people on board. The BEA’s final report (July 2012) found that the airspeed indications became temporarily inconsistent, probably because the Pitot probes were obstructed by ice crystals; the autopilot disconnected; the crew’s control inputs destabilised the flight path; and the crew did not identify that the aircraft had stalled, so no recovery inputs were made. The report made recommendations on training for manual handling at high altitude, on recognising an approach to stall, and on crew coordination. Leadership lesson: highly automated systems require people trained for the rare moment when automation hands back control.
4. Organisational silence and target pressure
Leaders often learn of serious problems last. People stay silent when they believe speaking up is futile or risky, when hierarchy is steep, or when the organisation rewards hitting targets more than telling the truth. Silence then looks like good news in the reports.
Targets are useful, but when a single number dominates — finance, throughput, waiting times — people adapt to meet the number, sometimes at the expense of the purpose it was meant to serve. Executives should pair every target with a measure of what could be harmed in pursuing it, and go to see the work themselves.
Key point: Pair every target with a measure of what it could damage.
Case study: Mid Staffordshire NHS Foundation Trust (UK). The public inquiry chaired by Robert Francis QC reported on 6 February 2013, making 290 recommendations. It described appalling and unnecessary suffering of many patients at Stafford Hospital between 2005 and 2009, and found that the Trust board had prioritised its finances and its pursuit of Foundation Trust status at the cost of acceptable standards of care, failing to listen to patients and staff. Warning signs existed in the wider system but were not acted on. The report called for a culture of openness, transparency and candour.
5. Tools for better decisions
The pre-mortem, described by psychologist Gary Klein in Harvard Business Review (2007), asks a team to imagine that a decision has already failed and to write down why. Because failure is assumed, people feel freer to voice doubts, and the team surfaces risks that optimism had hidden.
Other practical tools: ask for written views before discussion; give a respected person the role of devil’s advocate or a small red team the job of attacking the plan; set kill criteria in advance (what would make us stop?); and record the decision with its assumptions and dissent.
Finally, review decisions later — separating the quality of the decision from the luck of the outcome. Good decisions can turn out badly and bad ones can turn out well; only reviewing the process improves future judgement.
Key point: Set kill criteria before you start; it is much harder after you have invested.
The full lesson, case studies, scenarios and templates continue in the interactive course on this page.
Sources and references
- Report to the President (June 1986)
- Report, Volume I (August 2003)
- Final report on the accident on 1 June 2009 to the Airbus A330-203, flight AF 447 Rio de Janeiro–Paris (July 2012)
- Report of the Mid Staffordshire NHS Foundation Trust Public Inquiry (Francis Report, 6 February 2013)
- HSG48 Reducing error and influencing behaviour
- Gary Klein, “Performing a Project Premortem” (September 2007)
Executive education only, not an accredited qualification. Case facts are summarised from the official investigation reports cited; they are used to illustrate decision-making, not to assign individual blame.