Latest / Safe As Podcast / Ep 13: Do near misses increase risky decision making?

Transcript

Ep 13: Do near misses increase risky decision making? · Safe As Podcast

• 0:06
  Common logic and decades of safety practise tells us that
• 0:09
  reporting and investigating near misses makes us safer.
• 0:13
  It makes sense, right? If we learn from the almost
• 0:15
  disasters, we prevent the real ones.
• 0:17
  But what if that common wisdom, when applied in certain ways,
• 0:22
  actually creates a dangerous blind spot?
• 0:25
  Good day everyone. I'm Ben Hutchinson and this is
• 0:27
  Safe as a podcast dedicated to the thrifty analysis of safety,
• 0:31
  risk, and performance. Visit safetyinsights.org for
• 0:35
  more research. Today we're exploring a 2016
• 0:38
  paper from Dylan and Tinsley called Near Miss Events, Risk
• 0:42
  Messages and Decision Making, Publishing Environment, Systems
• 0:46
  and Decisions. They studied how near miss
• 0:49
  events and their reporting might actually increase our exposure
• 0:52
  to risk, rather than always improving learning and safety.
• 0:57
  What were their methods? Well, it used 2 experiments with
• 1:00
  participants. Experiment one was the cruise
• 1:03
  dilemma. In the first experiment,
• 1:05
  participants read a scenario about a planned cruise
• 1:08
  potentially being impacted by a hurricane.
• 1:11
  They were randomly put into one of three groups, the control
• 1:14
  group said. The control group was told that
• 1:16
  the cruisers can be diverted by bad weather, but they weren't
• 1:19
  given any info about past hurricanes in the area.
• 1:22
  Simple and neutral, there was the resilient Near Miss group.
• 1:26
  Here, participants were told about hurricane forecasts, but
• 1:29
  crucially, their last three trips were not diverted.
• 1:33
  This is where people perceive that disaster was successfully
• 1:36
  avoided, often leading them to discount future warnings.
• 1:40
  The third group, the Vulnerable Near Miss group, So a similar
• 1:43
  scenario, but this time a close friend had their crews diverted
• 1:47
  due to a storm. This scenario is where people
• 1:50
  perceive a disaster came very close and it encourages them to
• 1:53
  seek out future warnings and take mitigating actions.
• 1:57
  So what did they find? Well, in experiment one, when
• 2:01
  participants had information about a priori near miss, like a
• 2:04
  cruise almost been diverted, but they had no specific information
• 2:08
  about problems that almost occurred.
• 2:10
  So the resilient near miss group, they overwhelmingly chose
• 2:14
  to take the risky option to continue the cruise as planned.
• 2:17
  This was compared to both the control group and those in the
• 2:20
  vulnerable near miss condition. Interestingly, thinking the
• 2:24
  likelihood of problems was lower did lead people to choose the
• 2:27
  cruise, but the type of near misinformation didn't actually
• 2:31
  change their perception of how likely a problem was, they just
• 2:35
  felt differently about the risk. Also, risk seeking personalities
• 2:40
  were predictably more likely to choose riskier options, but this
• 2:45
  trait didn't interact with the near misinformation itself.
• 2:49
  Crucially, the type of near miss info did impact general
• 2:52
  optimism. Those with resilient near miss
• 2:55
  info seeing averted disaster as a positive, You know, we handled
• 2:59
  it. We're more optimistic and more
• 3:01
  likely to go on a cruise. What did they find in experiment
• 3:05
  too? This expanded on the resilient
• 3:07
  near miss data. Looking at how framing of
• 3:10
  uncertainty or danger influences decisions.
• 3:13
  Found that when near misinformation highlighted the
• 3:15
  likelihood of a negative event or its potential consequences,
• 3:19
  people were more likely to plan mitigation actions, especially
• 3:23
  compared to receiving no specific risk messaging.
• 3:26
  This suggests that the messaging matters.
• 3:29
  So what do we make of the findings?
• 3:31
  So these experiments strongly suggest that nearness events and
• 3:35
  how we report them may contribute to a state of
• 3:38
  heightened risk, the near failures or invertedly, A dull
• 3:42
  perception of risk, the near successes or how we narrowly
• 3:46
  created success. Perhaps our decisions can slowly
• 3:49
  drift towards the margins of safety over time.
• 3:52
  Why? Because some near misses
• 3:55
  actually dampen our feelings of vulnerability to harm, rather
• 3:59
  than sensitising us to them. The author's suggestion for
• 4:02
  emergency managers and safety leaders to overcome this
• 4:05
  negative impact is that when we communicate about a near miss,
• 4:09
  we should highlight how the event was almost a bad event,
• 4:12
  emphasising as a near failure rather than a near success.
• 4:16
  People with prior near miss info that highlights the averted
• 4:19
  disaster as a success are less likely to opt out of a
• 4:23
  potentially hazardous trip the near success.
• 4:26
  This isn't necessarily because they update their logical
• 4:28
  estimates of the likelihood there was no evidence of this,
• 4:32
  but rather they feel differently about the statistical risk,
• 4:35
  becoming more optimistic and less cautious.
• 4:39
  Ultimately, these findings are just to rethink how we learn
• 4:42
  from near misses, ensuring they truly sharpen our awareness of
• 4:45
  risk, not dull it Now. In fact, these authors found
• 4:49
  these results to be fairly stable across other
• 4:52
  organisations and roles also, so they have other research that
• 4:55
  confirms this. So this doesn't mean we
• 4:57
  shouldn't report, discuss or act on near miss events, but should
• 5:01
  consider how we frame them, particularly within the sphere
• 5:04
  of safety critical domains. There's also a lot more stuff to
• 5:07
  learn from daily work that doesn't rely on rare events, but
• 5:12
  that likely has its own caveats too.
• 5:14
  So limitations really. One of the core limitations is
• 5:18
  that they only looked at intended behaviour rather than
• 5:20
  actual behaviour. But in saying that, intended
• 5:24
  behaviour has been shown in other work to be a good
• 5:26
  predictor of actual behaviour. So in summary, according to this
• 5:32
  research, we should emphasise the near failure rather than the
• 5:36
  near success. That's it on Safe As I'm Ben
• 5:40
  Hutchinson. Please help share, rate and
• 5:42
  review, and check out Safety Insightsorg for more research.
• 5:47
  Finally, feel free to support Safe As by shouting a coffee
• 5:51
  link in the show notes.