27
Human Factors in Accident Investigation
ARE SOME PEOPLE ‘ ACCIDENT PRONE?
CHARACTERISTICS, CAUSES AND MITIGATION
The concept of ‘ accident proneness’ dates from World War 1 when it was found
that the statistical distribution of large numbers of accidents in munitions factories
did not fit the theoretical distribution ( a ‘ Poisson Distribution’) as expected – some
people had more accidents than they should have! The concept lost favour in the latter half of the 20th century as attention shifted towards the design of safety systems
and equipment in accordance with the focus of HF. More recently, there has been
renewed interest in accident proneness and several studies have shown that a measure of cognitive failure in daily l ife – ‘ The Cognitive Failures Questionnaire’ ( CFQ,
Broadbent et al. 1998) – can predict safety behaviour. The CFQ has questions about
common slips, lapses of attention and memory problems ( such as forgetting names)
in daily life. Most people report at least few such failures in daily life and some report
more than others. Larson et al. ( 1997) found a link between scores on the CFQ and
actual mishaps which concurs with other research that demonstrates that accidents
can result from distractibility, mental error and poor selective attention. Wallace and
Chen ( 2005) found that people with high scores on the CFQ were more likely to have
accidents due to lower attentiveness. In other words, accident proneness is nothing
more than the tendency towards cognitive failure in daily l ife – we are all accident
prone. To date, the evidence indicates that cognitive failure is what psychologists call
a ‘ trait’ – a stable characteristic that varies between people.
There is also evidence that accident proneness varies within people. Day et al.
( 2012) found that a link between high scores on the CFQ and having an accident was
mediated by psychological strain ( caused by exposure to psychosocial stress in the
workplace). People under stress are more susceptible to cognitive failures, and these
failures then cause accidents.
These findings suggest that accident proneness is both a trait ( a stable feature
of some individuals) and a state ( a condition that can affect all, if under sufficient
stress). Since adverse reactions to stress at work increase the propensity for accidents
to happen, psychosocial stressors in the workplace can be considered to be PSFs in
safety management and should be considered during safety investigations. The concept is more useful when the role of stress is considered, implying that a closer link
is needed between health monitoring and safety management. Since it is easier to
change working conditions than it is to change people, workplaces can be made safer
if they are designed to suit accident prone employees ( i.e. if they are designed to be
resilient when cognitive failures occur).
Case Study of Poor Human Factors Design –
Grounding of the Cruise Ship Royal Majesty
Dekker ( 2014) gives an interesting analysis of the circumstances leading to the
grounding of the cruise ship Royal Majesty in June 1995. The ship departed from
St. Georges, Bermuda, en route to Boston in the USA. The ship’s navigator carried
out the required navigational procedures, setting the ship’s autopilot to the correct
Human Factors in Accident Investigation
ARE SOME PEOPLE ‘ ACCIDENT PRONE?
CHARACTERISTICS, CAUSES AND MITIGATION
The concept of ‘ accident proneness’ dates from World War 1 when it was found
that the statistical distribution of large numbers of accidents in munitions factories
did not fit the theoretical distribution ( a ‘ Poisson Distribution’) as expected – some
people had more accidents than they should have! The concept lost favour in the latter half of the 20th century as attention shifted towards the design of safety systems
and equipment in accordance with the focus of HF. More recently, there has been
renewed interest in accident proneness and several studies have shown that a measure of cognitive failure in daily l ife – ‘ The Cognitive Failures Questionnaire’ ( CFQ,
Broadbent et al. 1998) – can predict safety behaviour. The CFQ has questions about
common slips, lapses of attention and memory problems ( such as forgetting names)
in daily life. Most people report at least few such failures in daily life and some report
more than others. Larson et al. ( 1997) found a link between scores on the CFQ and
actual mishaps which concurs with other research that demonstrates that accidents
can result from distractibility, mental error and poor selective attention. Wallace and
Chen ( 2005) found that people with high scores on the CFQ were more likely to have
accidents due to lower attentiveness. In other words, accident proneness is nothing
more than the tendency towards cognitive failure in daily l ife – we are all accident
prone. To date, the evidence indicates that cognitive failure is what psychologists call
a ‘ trait’ – a stable characteristic that varies between people.
There is also evidence that accident proneness varies within people. Day et al.
( 2012) found that a link between high scores on the CFQ and having an accident was
mediated by psychological strain ( caused by exposure to psychosocial stress in the
workplace). People under stress are more susceptible to cognitive failures, and these
failures then cause accidents.
These findings suggest that accident proneness is both a trait ( a stable feature
of some individuals) and a state ( a condition that can affect all, if under sufficient
stress). Since adverse reactions to stress at work increase the propensity for accidents
to happen, psychosocial stressors in the workplace can be considered to be PSFs in
safety management and should be considered during safety investigations. The concept is more useful when the role of stress is considered, implying that a closer link
is needed between health monitoring and safety management. Since it is easier to
change working conditions than it is to change people, workplaces can be made safer
if they are designed to suit accident prone employees ( i.e. if they are designed to be
resilient when cognitive failures occur).
Case Study of Poor Human Factors Design –
Grounding of the Cruise Ship Royal Majesty
Dekker ( 2014) gives an interesting analysis of the circumstances leading to the
grounding of the cruise ship Royal Majesty in June 1995. The ship departed from
St. Georges, Bermuda, en route to Boston in the USA. The ship’s navigator carried
out the required navigational procedures, setting the ship’s autopilot to the correct