Can near misses increase risky decision-making?
This week on Safe As, I spoke with Professor Robin Dillon-Merrill covering her work on how near misses can desensitise us to risk – challenging the rhetoric around the importance of reporting them.
Robin discussed how organisations routinely misinterpret near misses and mistake luck for system resilience.
Namely:
• There’s a difference between resilient vs vulnerable near misses – vulnerable (e.g. floodwater stopping an inch from a door) highlights vulnerability and triggers caution. But a resilient near miss (e.g. water two inches away purely to luck) doesn’t feel like danger, so isn’t interpreted as such.
• The issue of outcome bias in decisions – evaluating decisions, like a manager’s competence, is influenced by luck. Evaluators tend to rate people who experience near misses or lucky successes as more competent and promotable than those who experience failure, even when chance was the only difference and the underlying decision quality was identical.
• When people are presented with near miss info or minor warning signals, they may counterintuitively search for *less* additional info before making risky decisions
The Quickisode also explored whether fatigue contributes to inattentional blindness. The study from Li et al used eye-tracking tech, evaluating how mental time-on-task fatigue affects an excavator operator’s ability to spot peripheral hazards (people walking behind/to the sides of an excavator).
They found:
• A form of visual tunnelling occurs: >60 mins of continuous task, fatigued operators fixated their gaze less on the rear-view mirrors, and concentrated their gaze almost exclusively on the central display/forward facing.
• Missing people walking towards the rear of the excavator increased by over 40% after an hour, and 30% misses after just 36 mins.
• Fatigue doesn’t just merely slow mental processing and reaction times, but alters where an operator looks.
Check out Safe As YouTube or on Spotify/Apple. Please help share with your network if you find this work helpful 🙏
