ATPL Human Factors (CAA) common mistakes
ATPL Human Factors · Understand the traps and how to avoid them.
When a question describes a pilot who keeps flying despite growing uncertainty (weather, fuel, alternates), choosing the answer that describes the pilot as 'unaware' or 'lacking knowledge' rather than recognising a motivated, plan-continuation bias such as confirmation bias or plan continuation/ sunk-cost pressure.
Candidates memorise bias labels in isolation and match them to keywords, so a scenario without the word 'bias' gets answered with a generic competence deficit. Plan continuation is the most examinable bias because it explains why experienced pilots press on.
What to do instead
For any 'pilot continues despite...' stem, first ask 'what information did the pilot downplay, and what did they want to be true?' If the pilot wanted the original plan to work, answer plan continuation/goal seduction; if they only sought confirming data, answer confirmation bias. Practise sorting ten such vignettes into bias categories without looking at options.
Treating 'situational awareness' and 'workload' as synonyms, so a question about a pilot who is busy but still monitoring correctly is answered as 'loss of situational awareness', and a question about an overloaded but aware pilot is answered as 'poor SA'.
Both concepts appear in the same scenario stems and both degrade under high demand, so candidates collapse them into one idea and lose the distinction the examiner is testing.
What to do instead
Define them separately before answering: workload is the demand-capacity relationship; SA is the perception-comprehension-projection of what is happening. In each stem, locate the failure point (too much demand vs wrong mental model) and answer only that. Write one sentence each for SA levels 1, 2 and 3 and check which level the stem describes.
Answering fatigue questions with the general rule 'rest more' instead of the specific mechanism asked about, e.g. misidentifying the circadian low points, confusing sleep debt with circadian rhythm, or attributing a night-time performance drop to cumulative fatigue when the stem describes the circadian trough.
Fatigue, sleep and circadian content is taught together and the mechanisms overlap in real life, so the distinctions blur under time pressure; candidates also rely on lay beliefs about 'catching up on sleep'.
What to do instead
Build a three-column table: circadian (time-of-day, two lows, hard to override), sleep debt (cumulative, partly repayable), and acute fatigue (time-on-task, recoverable with a break). For each question, identify which column the stem's evidence fits before reading the options.
In CRM/communication questions, selecting the answer that describes what the captain should have 'known' or 'told' rather than the behaviour that would have closed the loop, such as a readback, a challenge-and-response, or an explicit inquiry by a junior crew member.
Candidates read the scenario from the captain's perspective and answer with authority-based reasoning, which feels natural but contradicts the assertiveness and closed-loop communication content the exam rewards.
What to do instead
When a scenario contains a junior crew member with information, ask 'what phrase would have transferred that information unambiguously?' Prefer answers containing readback, standard phraseology, inquiry or a stated intention. Reject answers that rely on assumption, seniority or 'should have noticed'.
Misclassifying error types in error-model questions, for example labelling a slip as a mistake, a violation as an error, or a lapse as a decision error, because the candidate matches on outcome severity rather than on the intention and the stage of the action.
Slips, lapses, mistakes and violations are defined by intention and execution, but exam stems describe consequences, which pulls attention towards severity. Candidates also forget that violations are deliberate deviations, not unintentional errors.
What to do instead
Use a two-question decision tree: was the action intended? (no = slip/lapse; yes = mistake or violation). Was the intention correct but execution wrong? (slip = observable action, lapse = omitted step). Was a rule knowingly broken? (violation). Apply this to every error question before reading options.
Choosing 'increase automation level' or 'disengage automation' as the fix for an automation-related problem without checking whether the stem describes a mode awareness problem, a complacency problem, or a skill-degradation problem.
Automation is taught as a list of advantages and disadvantages, so candidates answer with a general attitude rather than diagnosing the specific failure. Mode confusion in particular is often misread as a system malfunction.
What to do instead
For automation stems, first name the failure: mode confusion (what is the aircraft doing vs what the crew thinks), complacency (monitoring dropped because reliability is high), or skill degradation (manual handling eroded). The fix must match: for mode confusion, verify mode annunciation and fly raw data if needed; for complacency, active monitoring and callouts.
Answering decision-making questions with the 'best' option in hindsight rather than the option that reflects a recognised decision model, such as the difference between analytical (DECIDE-type) and naturalistic/recognition-primed decisions, or between a good decision and a good outcome.
Candidates conflate outcome quality with decision quality and pick the option that produced the safest result, even where the stem asks about the process.
What to do instead
Read the question stem for the word 'process', 'model' or 'why'. If it asks about process, choose the answer describing the decision strategy, not the result. Learn the distinction: good decision with bad outcome is still a good decision. Practise separating process from outcome in ten short scenarios.
Confusing the physiological mechanisms in the physiology domain, for example attributing hypoxia symptoms to hyperventilation, mixing up the stages of spatial disorientation (especially somatogyral vs somatogravic illusions), or mislabeling the effects of barometric pressure change on the middle ear and sinuses.
Many of these conditions share symptoms (dizziness, tingling, visual change), and candidates learn symptom lists without the associated cause or flight phase, so similar-sounding options are hard to discriminate.
What to do instead
Attach each condition to a trigger and a flight phase: somatogravic = acceleration during takeoff/go-around; somatogyral = sustained rotation after a spin or rapid turn; hyperventilation = anxiety with tingling and carpopedal spasm, relieved by slowing breathing; hypoxia = decreasing performance with altitude, relieved by oxygen. Build cue cards linking trigger to illusion to correct response.
In perception and attention questions, treating attention as a single resource and answering 'the pilot was inattentive' where the stem describes a failure of divided attention, attentional tunnelling, or change blindness, or confusing bottom-up capture with top-down expectation.
Attention models are abstract and candidates reduce them to a single 'concentration' idea, so they cannot distinguish the specific mechanism the examiner is describing.
What to do instead
Learn four specific labels and their signatures: selective attention (filtering), divided attention (two tasks competing), attentional tunnelling (narrowing under stress or high demand), and change blindness (failure to notice a change in a fixated scene). Match the stem's evidence to the signature, especially the presence of high stress or a head-down task.
Treating stress and arousal as a single linear effect, so a stem about a pilot performing poorly under high arousal is answered as 'too little arousal' or the Yerkes-Dodson curve is applied without noting that the optimum shifts with task complexity.
The inverted-U is often memorised as a picture without the qualifier that complex tasks have a lower optimal arousal level, and candidates confuse acute stress with chronic stress responses.
What to do instead
State the rule explicitly: performance peaks at moderate arousal; the optimum is lower for complex/new tasks and higher for simple/well-learned ones. When a stem describes a degraded performance in a demanding task, check whether arousal is above or below the optimum for that task complexity before choosing. Also separate acute (fight-or-flight, immediate) from chronic (cumulative, health effects) stress responses.