Breaking Down Biases

Cognitive biases can subtly, even subconsciously, influence decision-making in aviation with serious consequences. Understanding these biases and employing strategies to mitigate them is essential for pilots at all experience levels.

Let’s consider two biases that are among the most concerning in aviation: confirmation bias and anchoring bias.

Confirmation bias

On January 21, 2019, CommutAir Flight 4933 overran the runway at Presque Isle International Airport in Maine while landing in snowy, low-visibility conditions. The crew was attempting an ILS approach but didn’t have a clear visual of the runway.

Despite this, they continued their descent, influenced by their expectation that the runway would appear. This led them to assign false truth to vague visual cues as confirmation of the runway’s position, causing the aircraft to land on a snow-covered grass area to the right of the runway.

The NTSB investigation identified confirmation bias as a key factor in the incident. Contributing factors included fatigue in the first officer, poor decision-making by the captain, and a previously reported misalignment of the ILS localizer that had not been addressed.

Most people have heard of confirmation bias, as it’s the most common type of bias. This is when the mind seeks out or interprets information to support a preexisting belief (or hope) while ignoring contradictory data.

The risk here is that a pilot might make flawed decisions simply because s/he focuses only on partial data that implies the situation is something other than what it is in reality. Fear-motivated confirmation bias is especially dangerous because, for comfort, we can talk ourselves into rationalizing a false conclusion that favors a desirable outcome. Unfortunately, objective reality doesn’t care that we’re in denial.

Here's another example of confirmation bias at a stressful moment during a flight.

A pilot is on an instrument approach in deteriorating weather conditions. Based on the pre-flight briefing and earlier weather reports, the pilot expects visibility to be sufficient to complete the approach safely. However, updated ATIS reports inform the pilot that visibility has dropped below minimums.  

Instead of reevaluating the decision to land based on the new ATIS information, the pilot disregards it, focusing instead on the belief that visibility will improve to match the original prediction that the visibility won’t be too bad to land for another 30 minutes. Perhaps the pilot concludes that the current low visibility is just a transient patch of cloudiness.

The pilot has just selectively interpreted objective data and then filled in the rest of the picture with hopes and assumptions that led to a plausible scenario. Unfortunately, while it might have been plausible, it didn’t reflect reality, causing the pilot to have a bad day.

Anchoring bias

For you CFIs out there who remember your FOI study, anchoring bias is related to primacy. While they are distinct phenomena, primacy can be a contributing factor to causing an anchoring bias.

Anchoring bias is present when a person relies too heavily on the first piece of information received, hence the term “anchor.” The mind becomes fixed on that first piece of information and is reluctant to change decisions even when new information contradicts the reports received earlier.

The primacy effect, which states that people tend to recall the first and last things they heard best, can somewhat reinforce an anchoring bias situation.

Here’s an example.

A pilot is planning a fuel stop on a long cross-country flight. The pilot checks fuel prices at different airports and finds an excellent price at “Airport A.”

It’s an airfield with which the pilot is familiar, which generates some confidence. The pilot decides to go there, thus becoming the pilot’s “anchor.”

During the flight, conditions change, and strong headwinds cause increased fuel burn. The aircraft’s remaining fuel is now insufficient to reach Airport A and land with legal reserves.

Despite this, the pilot continues toward Airport A, reluctant to deviate because s/he is anchored to the idea of refueling there due to the earlier perceived benefit (the low fuel price) and the confidence of landing somewhere familiar.

We need not conclude this story to understand that even a best-case outcome involves a lot of stress.

Mitigation strategies

To combat confirmation bias, pilots should develop the habit of questioning their assumptions and identifying feelings of discomfort that might be distorting important facts or removing them from consideration. This requires a degree of humility and courage on the part of the pilot in command. A pilot must be willing to admit that their assumptions were incorrect or that a situation involves danger.

Similarly, to mitigate anchoring bias, pilots should treat the information they first receive as provisional and remain open to revising their plans based on updates and real-time data received during the flight, especially when new data evokes a negative emotion. Being confident about several options instead of just one familiar action or location makes it easier to adapt to whatever situation develops inside or outside the cockpit.

For both:

• Read up on incidents in which cognitive biases were contributing factors and discuss scenarios with other pilots in which biases could arise.
• Adhere to standardized procedures, such as checklists and decision-making models like DECIDE, to better ensure your decisions are based upon accurate and complete information.
• Engage ATC for added perspective.
• Regularly reassess the situation, especially during extended legs or complex flights, to update decisions in alignment with current conditions and data.

Know Before You Go