Cognitive Biases in Decision-Making
A cognitive bias is a systematic, predictable error in the way people perceive information, judge probability or choose between options. In aviation, biases such as confirmation, expectation and plan continuation bias lead pilots to keep a failing plan and to misread the evidence against it.
A cognitive bias is a systematic error in the way people take in information, estimate probability or choose between options. Biases are not random mistakes and not signs of incompetence: they are the by-products of mental shortcuts that work well most of the time, and they affect experienced pilots as much as students. Because they are predictable, they can be anticipated and designed against.
In flight, biases do most harm at the moment a plan should change. The evidence that the weather is worse than forecast, the approach is not stabilising or the fuel will not be enough is usually available; biases make it less visible, less credible or less urgent than it is. Accident reports return to the same few patterns, plan continuation, get-there-itis, scud running and descent below minima, which the FAA groups as operational pitfalls. The EASA Human Performance and Limitations (040) syllabus and the FAA knowledge tests both examine them.
What cognitive biases are
People judge and decide with heuristics, rules of thumb that save effort. The psychologists Amos Tversky and Daniel Kahneman showed that these shortcuts produce consistent, measurable errors, and that the errors follow the heuristic rather than the individual. The same biases therefore appear in every crew room.
In the skill, rule and knowledge framework, biases matter most at the knowledge-based level, when no procedure fits and the pilot must reason from a mental model. Human factors texts list the errors of knowledge-based behaviour as incomplete or inaccurate mental models, overconfidence, lack of situational awareness, confirmation bias and frequency bias. Stress makes things worse: among its cognitive effects are fixation, regression to old habits and a compulsive search for information that confirms a decision already made. Biases cannot be removed by knowing about them. They are managed by procedures, criteria set in advance and a crew that is free to challenge.
Confirmation and expectation bias
Confirmation bias is the tendency to seek, notice and give weight to information that supports a view already held, while ignoring or explaining away evidence against it. A pilot who consults four weather sources before a marginal flight but stops as soon as one forecast supports departure is showing it. So is a crew that, having diagnosed a fault, reads every later indication as proof of the diagnosis. The recognised defence is to look deliberately for information that would falsify the hypothesis rather than confirm it. For weather decisions, deciding in advance which source governs, and reading it last rather than first, takes the choice out of the bias's hands.
Expectation bias acts on perception itself. The brain builds what it sees and hears from fragments of the signal plus context and expectation, and under a weak signal or high workload the expectation wins. A pilot cleared to 3,000 ft on every previous arrival may hear "descend to 3,000 feet" when the controller said 5,000, and read it back with conviction; a pilot waiting for descent may accept a clearance meant for a similar callsign. The readback does not catch the error, because the pilot reads back what he believes he heard. Defences are full callsigns, listening to the whole message before acting, and questioning any clearance that does not fit the situation.
Availability, frequency and framing
The availability heuristic judges how likely an event is by how easily examples come to mind. Well-publicised events, such as fatal airliner accidents, are overestimated, while less salient ones, such as airprox incidents and non-fatal accidents, are greatly underestimated. People also tend to overestimate the frequency of rare but favourable outcomes, the logic of the lottery ticket, which in flight becomes the belief that the gap in the weather will be there. Frequency bias is the related tendency to call to mind scenarios met often before and apply them where they do not fit. Saliency draws attention to cues that are loud, bright, recent or centrally placed, so that vital but quieter information is missed.
The framing effect is the finding that the same choice is decided differently depending on how it is presented. Offered two gains, people prefer the certain one. Offered two losses, a certain loss or a gamble on a larger one, people tend to choose the gamble, even when its expected loss is greater. Turning back is a certain loss: a missed appointment, disappointed passengers, a night away. Pressing on is a gamble that usually pays off. Framed that way, the risky choice feels natural, and human factors texts name this bias as a source of press-on-itis.
Denial protects the decision from the evidence. Its typical forms each have a voice: procrastination ("I'll continue a little longer and decide later"), rationalisation ("it'll all work out"), hope ("it's bound to clear on the other side"), refusal to admit ("it can't happen to me"), status ("I'm not going to be beaten") and refusal to review ("it worked last time"). Rationalisation is especially insidious because it leaves the pilot feeling reasonable: a pilot who landed from below minima and explains that "the reported visibility there is always pessimistic" has removed the event from any lesson.
Plan continuation bias
Plan continuation bias is the tendency to persist with an original plan despite mounting evidence that it should change. It is the most common form of tactical decision error in airline accidents: a NASA review of the 37 crew-involved accidents studied by the NTSB in 1994 found that almost three quarters of the tactical decision errors were decisions to continue with the original plan despite cues suggesting a change. It strengthens as the goal approaches: the effort already invested, fatigue and the pull of the destination make abandoning the plan feel more costly than continuing. It works by making each new piece of bad news look like an acceptable variation, so that no single moment feels like the moment to stop.
A single pilot 90 minutes into a two-hour flight who finds the destination below personal minima, yet presses on rather than diverting to a suitable aerodrome 20 NM behind, is showing it. So is a crew that continues an approach into worsening weather because the runway is close. The defence is a hard decision gate chosen before the pressure arrives: diversion criteria briefed before descent, a fuel figure at which the alternate becomes mandatory, stabilised approach criteria that trigger a go-around without discussion, and published minima treated as a limit (see stabilised approach).
Get-there-itis and external pressure
Get-there-itis, also called get-home-itis or press-on-itis, is a fixation on reaching the destination that clouds judgement about weather, fuel, fatigue or the aircraft. Its fuel is external pressure: a meeting, a wedding, passengers who want to continue, the last day of a trip. In commercial flying the same force appears as commercial pressure or operational pressure, the expectation that flights depart and arrive on time, that fuel and diversions cost money and that delays are noticed.
The pressure is more dangerous in single-pilot commercial operations than in a two-crew aircraft, because there is no second crew member to challenge a flawed plan; risky shift, the group tendency towards bolder decisions, cannot apply to a lone pilot, so the pressure he feels is his own. The defence is pre-commitment. Decisions made calmly on the ground, written personal minimums, a cancellation time, diversion criteria and a warning to passengers that the weather decides, are far harder to renegotiate than decisions made with the destination in sight. Operators help with a visible no-fault delay and diversion policy.
VFR into IMC and scud running
Continued VFR into IMC is one of the deadliest outcomes of these biases. A VFR pilot who presses on into deteriorating weather eventually loses visual reference. In a University of Illinois study, often cited as "178 seconds to live", pilots without instrument training placed in simulated instrument conditions all lost control, entering a spiral or another dangerous attitude after 178 seconds on average. The taught response to inadvertent IMC is an immediate, gentle 180° turn back towards visual conditions by reference to the instruments.
Scud running usually comes first: trying to keep the ground in sight by flying ever lower under a lowering cloud base and in poor visibility. It pushes pilot and aircraft to their limits, brings them close to towers, wires and rising ground, and often ends in inadvertent entry into cloud. Mountain obscuration is especially dangerous, because the valleys may be clear while the passes are hidden. Treating an obscured pass as closed, and turning back while there is still room to turn, is the answer.
Duck-under syndrome
Duck-under syndrome is the instrument pilot's version of the same trap: descending below the minimum descent altitude or decision altitude without the required visual references, because the runway is expected to appear at any moment or because the pilot is determined to land. The minima exist because below them the approach no longer guarantees obstacle clearance; a pilot who "has a look" below them has left the protected procedure. The FAA lists duck-under syndrome as an operational pitfall, and descent below the minimum en route altitude as a separate one.
Operational pitfalls
The FAA's list of operational pitfalls brings these patterns together. It includes:
- peer pressure, poor decisions driven by the opinions of peers;
- mindset, an inability to recognise and cope with changes from what was expected;
- get-there-itis, duck-under syndrome, scud running and continuing VFR into IMC;
- getting behind the aircraft and loss of positional or situational awareness;
- operating without adequate fuel reserves;
- descent below the minimum en route altitude and flying outside the envelope;
- neglect of flight planning, preflight inspections and checklists.
Most of them are the same biases seen from the cockpit: an expectation that the plan will work, reinforced by pressure to complete it.
Countering bias
Because biases operate below awareness, the defences are mostly structural:
- Set the gates on the ground. Personal minimums, diversion criteria, fuel figures and go-around criteria decided before departure cannot be eroded one small step at a time in flight.
- Seek disconfirming evidence. Ask what would make the plan unsafe, and look for it.
- Choose the governing source in advance, whether a particular forecast, the approach minima or a fuel figure.
- Say the plan aloud and invite the other pilot, or ATC, to challenge it; in single-pilot flying, brief it aloud anyway.
- Review at fixed points. Structured models such as DODAR and FOR-DEC end with a review or check step so that a decision which has quietly stopped being valid is caught (see aeronautical decision-making).
- Reframe the choice. A diversion is part of the plan, not a failure of it.
Exam tip: in EASA questions, the bias towards the risky loss when choosing between two losses explains press-on-itis; confirmation bias is avoided by looking for information that would falsify the hypothesis; and risky shift is a group phenomenon that cannot explain a lone pilot's decision.
Frequently asked questions
What is plan continuation bias?
Plan continuation bias is the unconscious tendency to keep following the original plan even as the evidence shows it should be changed. It is strongest near the end of a flight, when the destination is close and effort has already been invested, and it makes each new piece of bad news look like a minor variation rather than a reason to stop. Its main defence is a decision gate, such as diversion criteria, fixed before the pressure arrives.
What is get-there-itis?
Get-there-itis, also called get-home-itis or press-on-itis, is a fixation on reaching the destination that clouds judgement about weather, fuel, fatigue or the aircraft's condition. It is fed by appointments, passengers, schedules and commercial pressure, and it is one of the FAA's operational pitfalls. The defence is pre-commitment. Set personal minimums, alternates and a cancellation time before departure, and tell the passengers in advance that the weather decides.
What is scud running?
Scud running is trying to stay in visual contact with the ground by flying lower and lower under a lowering cloud base and in poor visibility. It pushes both pilot and aircraft to their limits, brings the aircraft close to towers, wires and rising ground, and often ends in inadvertent flight into cloud. The FAA lists it among its operational pitfalls. Turning back or landing early while there is still room to do so is the safe answer.
What is duck-under syndrome?
Duck-under syndrome is descending below the minimum descent altitude or decision altitude of an instrument approach without the required visual references, usually because the runway is expected to appear at any moment or the pilot is determined to land. It gives up the obstacle clearance on which the approach was designed, and the FAA lists it among its operational pitfalls. Published minima are a limit to respect, not a target to negotiate.
How can pilots counter confirmation bias?
The best defence against confirmation bias is to look deliberately for information that would prove the current plan wrong rather than information that supports it. In practice this means deciding in advance which weather source or indication will govern the decision, asking what would make the plan unsafe, inviting the other pilot to challenge, and reviewing the decision at fixed points instead of only when something forces a rethink.
Test yourself on Cognitive Biases in Decision-Making
The v1prep banks cover this topic in Human Performance and Limitations (040), with a worked explanation for every answer. EASA ATPL, PPL, IR and CPL, the FAA written tests and A320/B737 type ratings.
Start practising →Sources and further reading
- A. Tversky and D. Kahneman, Judgment under Uncertainty, Heuristics and Biases, Science 185, 1974
- A. Tversky and D. Kahneman, The Framing of Decisions and the Psychology of Choice, Science 211, 1981
- FAA Pilot's Handbook of Aeronautical Knowledge (FAA-H-8083-25), Chapter 2, Aeronautical Decision-Making (operational pitfalls)
- FAA Advisory Circular 60-22, Aeronautical Decision Making
- UK CAA CAP 737, Flight-crew human factors handbook
- SKYbrary, Continuation Bias
Library articles are written for study and exam preparation. They do not replace your aircraft's approved documentation, your operator's procedures or the regulations themselves.