Cognitive psychology research suggests pilots could be learning the wrong lessons from close-calls

New research provides evidence that two common cognitive biases could impact pilot’s perception of past events in ways that adversely affect how they make future decisions.

The findings, which appear in the journal Applied Cognitive Psychology, indicate that pilots could be learning the wrong lessons from close-calls — thanks in part to an error in thinking known as an outcome bias.

“I have been involved in aviation for many years; one of my key research areas has been the role cognitive biases play in pilot decision making. In aviation, when a poor decision is made, the consequence can be catastrophic, therefore having a greater understanding of what leads to poor decisions is an important step to improve aviation safety,” said study author Stephen Walmsley, who received a PhD in Aviation from Massey University.

The researchers were particularly interested in the often fatal mistake known as flying “VFR into IMC” — when pilots operating under visual flight rules inadvertently fly into low-visibility conditions that require instrument flight.

“Learning from previous poor decisions is imperative to avoid future mistake. However, what if our perception of past events is not a true reflection of what happened? This study explored cognitive biases that can influence our ability to learn from past events,” Walmsley told PsyPost.

“Consistent with other professional fields, pilots were influenced by outcome and hindsight bias. Of particular interest and concern was that ‘close-call’ events were treated similar to positive/safe outcomes.”

“Although the eventual outcome for a close call is the same as for a positive outcome, considerably more luck may be required in the close‐call situation to achieve that outcome. This may limit the learning opportunity from close-call events and reinforce risky behaviour,” Walmsley said.

To examine the potential impact of outcome bias, the researchers had 142 pilots read several fictional scenarios in which non-instrument rated pilots had taken off into questionable weather.

The beginning of each scenario was the same for each participant, but the researchers manipulated the outcome of the flight. In some cases, the flight was conducted without incident and the pilot landed safely. In close-calls, the pilot inadvertently entered instrument meteorological conditions but was able to safely turn around to regain visual conditions. In other cases, however, the pilot inadvertently entered instrument meteorological conditions and then crashed.

After reading each scenario, the participants were asked to rate the decision-making ability of the pilots and how much risk the pilots took.

The researchers found that the outcome of the scenarios had a significant impact on how the participants viewed decision-making and risk. Pilots who did not encounter any incidents were viewed as having better decision-making abilities and taking less risk than those who crashed — even though both hypothetical pilots took off into the same conditions.

Worryingly, participants treated pilots who had close-calls very similarly to pilots who did not encounter any incidents.

To examine the potential impact of hindsight bias, the researchers had another 62 pilots read about three planned cross-country flights which had yet to take place and then state their confidence regarding whether the flight would be safe/uneventful, require the pilot to turn around after encountering weather conditions, or result in a crash.

Ten days later, the participants were presented with the same three flights. This time, however, the participants were told the outcome of the flight. They were then asked to recall their initial predictions.

The participants tended to demonstrate hindsight bias for flights that ended safely and flights that ended in a crash. In other words, the participants believed they had originally assigned a higher probability to these outcomes than they actually did.

“Aviation is very safe, especially when compared to other modes of transport. Aircraft accidents are rare and when they do happen involve a range of factors. The cognitive biases highlighted in this study are unlikely by themselves to result in an accident, but can lead a pilot one step closer,” Walmsley explained.

“Caution needs to be applied when generalising these findings to the wider aviation population. The study participants primarily operated smaller aircraft with limited flight experiences.”

The study, “Understanding the past: Investigating the role of availability, outcome, and hindsight bias and close calls in visual pilots’ weather‐related decision making“, was authored by Stephen Walmsley and Andrew Gilbey.