Why the FAA uses CogScreen-AE: the science behind pilot cognitive assessment

TL;DR: The FAA commissioned CogScreen-AE in the late 1980s because existing neuropsychological tools were not built for pilots, and a general intelligence test cannot tell you whether someone can simultaneously track an ILS, manage a radio failure, and hold a clearance in working memory. The research connecting CogScreen performance to actual flight performance is real, though not without legitimate critics. Understanding the science helps you approach the test as a measurement instrument rather than an obstacle, and that shift in framing tends to produce better outcomes.

The frustration is legitimate
Where CogScreen came from
What it actually measures, and why those things matter
The research connecting scores to flight performance
The honest part: where the debate continues
What this means for how you approach the test
Further reading on faacogscreen.com

The frustration is legitimate {#frustration-is-legitimate}

If you have spent any time in pilot forums after receiving a CogScreen referral, you have seen the range. The confused posts from pilots who cannot understand why memorizing digits backward has anything to do with flying a 737. The angry posts from experienced aviators with tens of thousands of hours who are baffled that a computer test might determine whether they keep their certificate. The anxious posts from pilots who failed and feel the test has no relationship to their actual performance in the cockpit.

That frustration is not irrational. The CogScreen-AE is a strange-looking test if you encounter it without context. The format is abstract. The connection to flying is not immediately obvious. And the stakes attached to it are serious enough that the lack of obvious face validity feels like an insult.

So let's actually talk about where it came from, what it is designed to measure, and what the evidence says, including the parts where reasonable people disagree.

Where CogScreen came from {#where-cogscreen-came-from}

Thirty years ago, work was initiated to develop CogScreen-Aeromedical Edition as a computer-administered and scored cognitive screening instrument. The instrument was originally designed for the US Federal Aviation Administration to rapidly assess deficits or changes in attention, immediate and short-term memory, visual perceptual functions, sequencing functions, logical problem solving, calculation skills, reaction time, simultaneous information processing abilities, and executive functions.

Dr. Gary Kay, then Director of the Neuropsychology Division at Georgetown University School of Medicine, led that development effort under FAA contract, producing the professional manual in 1995.

The context matters. Before CogScreen, the tools available for evaluating pilot cognition were either standard neuropsychological batteries developed for clinical populations (people with brain injuries, dementia, psychiatric conditions) or general intelligence assessments designed for employee selection. Neither was built around the specific cognitive demands of operating an aircraft. A general IQ test might tell you that a pilot is in the 85th percentile of verbal reasoning, but it cannot tell you much about their ability to hold an ATC clearance in working memory while simultaneously cross-checking instruments and monitoring for traffic.

The FAA needed something purpose-built. CogScreen-AE was the answer. CogScreen-AE is a computer-based assessment tool commissioned by the FAA and developed by Kay (1995) to evaluate cognitive function in pilots. The hour-long assessment consists of 13 tasks that measure various cognitive abilities, including visual perception, attention, memory, decision-making, and motor skills.

What it actually measures, and why those things matter {#what-it-measures}

The 13 subtests on CogScreen-AE are not arbitrary. Each targets a cognitive function that has documented relevance to flight operations. Let me connect a few of them to what actually happens in a cockpit.

Working memory (Backward Digit Span, Auditory Sequence Comparison): When ATC gives you a complex clearance, you hold it in working memory while you read it back, cross-check your instruments, and write it down. Working memory is the cognitive workspace where that happens. Impairments in working memory are associated with errors in reading back ATC messages, particularly those containing longer digit strings.

Divided attention (Divided Attention Test, Dual Task): Flying IFR in actual IMC requires you to scan instruments, monitor the radio, track your position, and anticipate what comes next, all simultaneously. The divided attention subtests measure how well you manage competing demands on cognitive resources. This is not an abstract academic construct. It is a direct proxy for what managing cockpit workload requires.

Processing speed (Symbol Digit Coding, Pathfinder): Faster processing is not better for its own sake. In aviation, processing speed determines whether you can absorb incoming information (a traffic advisory, an amended clearance, a new weather report) quickly enough to act on it before the situation changes. Degraded processing speed creates lag between perception and response, and in a dynamic environment, that lag matters.

Spatial reasoning (Manikin, Figure Analysis): Instrument flying requires mental rotation, the ability to translate what you see on a display into a three-dimensional understanding of your aircraft's position in space. The Manikin subtest is a direct measure of this capacity.

Executive function (Pathfinder, Shifting Attention Test): Sequencing, cognitive flexibility, the ability to switch between tasks without losing track of where you were. These are the functions that allow a pilot to manage an unexpected systems failure while continuing to fly the aircraft and communicate with ATC.

If you want more detail on how each subtest maps to specific cognitive functions, the CogScreen-AE Subtests Guide on this site covers all 13 in depth.

The research connecting scores to flight performance {#research-connecting-scores}

The question that matters most is whether CogScreen scores actually predict anything about how pilots perform. The honest answer is: yes, with meaningful caveats.

Taylor et al. (2000) found that the speed and working memory factor was the strongest predictor of overall simulated flight performance. That study used a flight simulator to evaluate general aviation pilots and correlated their simulator performance against their CogScreen scores. The connection was statistically significant.

More specifically, Taylor and colleagues found that cognitive flexibility scores were significantly associated with four of the five key performance indicators during simulated cross-country flight, which included maintaining a prescribed course, dialing communication frequencies, traffic avoidance, and approach stability.

Those are not peripheral tasks. Maintaining course, managing communications, avoiding traffic, and stabilizing approaches are the core of what pilots do.

Military validation has also supported the instrument's utility. CogScreen-AE is already validated and used worldwide for detection of neurologic complaints in pilots, and results further support its use for screening military aviators to determine the need for comprehensive evaluation. When used as a screening test, a result within the normal range is sensitive enough to clear pilots from the need for a comprehensive neurocognitive evaluation before returning to duty.

The Van Benthem and Herdman research, which appeared after the Taylor et al. work, further developed the relationship between specific CogScreen factors and aviation performance outcomes. Their work helped establish the Taylor Aviation Factors framework that underlies the scored output you receive from a CogScreen evaluation.

The broader human factors context is also relevant. An analysis of all pilot-related accidents in the US from 1990 to 2000 indicated that about 80% were associated with skill-based errors, roughly 30% were associated with decision errors, and between 5% and 20% were related to violations and perceptual errors. Aviator proficiency and cognition accounts for almost all pilot-related accidents.

The FAA's interest in measuring pilot cognition is not bureaucratic excess. Cognitive function, specifically attention, memory, processing speed, and decision-making, is implicated in the overwhelming majority of aviation accidents. The question was never whether to care about it. The question was how to measure it consistently and fairly.

The honest part: where the debate continues {#honest-part-debate}

I mentioned that the frustration expressed in pilot forums is not irrational. Here is what that frustration is responding to.

CogScreen-AE has shown some association with line-check performance of airline pilots (Kay, 1995). Similarly, Taylor and colleagues and Van Benthem and Herdman found relationships between some CogScreen-AE subtests and flight performance, yet its ability to identify pilots who are definitively at risk for a serious incident or accident due to neurocognitive impairment has not been validated.

That last sentence is the crux of the legitimate criticism. The test can identify cognitive patterns associated with lower performance on certain simulated tasks. It has not been validated as a predictor of real-world accident risk in the way that, say, cardiovascular testing can predict cardiac events. The gap between "associated with lower simulator performance" and "predicts who will have an accident" is not a small one.

Some experienced medical examiners have raised concerns about the normative data, noting that it may not reflect modern pilot populations or modern simulators. The observation that the test may disadvantage pilots who are less familiar with computer interfaces is worth taking seriously, particularly for older aviators who did not grow up with the format. This is part of why preparation, specifically format familiarization and reducing interface anxiety before you sit down, matters. You are not gaming the test. You are ensuring that the interface itself does not suppress your actual performance.

The research base is also thinner than the test's widespread clinical use might suggest. Most of the foundational validation work appeared before 2000. That does not make the test invalid, but it does mean the science is not as settled as the institutional weight behind it might imply.

None of this changes the current reality: the FAA uses CogScreen-AE, HIMS AMEs rely on it, and neuropsychologists interpret it as part of a broader evaluation. Knowing where the science is strong and where it has limits helps you understand the process more clearly. It does not change the preparation strategy.

What this means for how you approach the test {#what-this-means}

Understanding the science behind CogScreen-AE changes two things that matter practically.

First, it reframes the test. The subtests are not arbitrary puzzles designed to trip pilots up. They are measurements of cognitive functions that research has linked to cockpit performance. When you work through a Divided Attention session in PilotPrep, you are training the same divided resource management that you use every time you fly in busy airspace. That connection is real. The test was built around it.

Second, it clarifies what good preparation looks like. The concern sometimes expressed in forums is that preparing for CogScreen means memorizing answers or gaming a test. That is not how any of this works. CogScreen-AE uses procedurally generated content. There are no answers to memorize. What preparation does is reduce the cognitive overhead caused by format unfamiliarity and test anxiety, so that on evaluation day, your scores reflect your actual cognitive capacity rather than how flustered you were by an interface you had never seen before.

For a detailed walkthrough of how scores are interpreted, including what pilot-normed percentiles actually mean and how the Taylor Aviation Factors are calculated, see Understanding Your CogScreen-AE Scores.

If your evaluation context involves an ADHD history, the FAA ADHD Disposition Table explained covers the Fast Track and Standard Track pathways and how CogScreen fits into each.

References

Dr. Jordan "Coach" Keller is an AI educator employed by PilotPrep LLC, created to help pilots navigate CogScreen-AE preparation and FAA medical certification. This post is educational and does not constitute medical advice.

Disclaimer: CogScreen-AE is administered and interpreted by licensed neuropsychologists. PilotPrep is a preparation and familiarization tool, not the actual CogScreen-AE test. Always consult a qualified HIMS AME and neuropsychologist for guidance specific to your case.

Why the FAA Uses CogScreen-AE: The Science Behind Pilot Cognitive Assessment