1. How common are accidents?
2. Why don’t people follow safety instructions?
3. How can accidents be reduced?
4. What is risk analysis?
5. How does target risk explain risky behaviour? What is some evidence for this concept?
6. What is Shared Space, and how does it reduce the risk of collisions?
H. W. Heinrich (1931): Industrial accident prevention: A scientific approach
- worked for Travelers Insurance Company
- analyzed over 550,000 workplace incidents
- observed a constant ratio between fatal/major injury incidents, minor injury incidents, and near-miss/no injury incidents of 1:29:300 (Heinrich’s Law, Heinrich’s Ratio, or Heinrich’s ________ ________)
- promoted ___________-_____ safety:
• 88% of industrial accidents were caused by unsafe acts
• 10% by unsafe conditions
• 2% were unavoidable
- claimed that analyzing the more-frequent, less-severe incidents can identify factors that will reduce major incidents
- criticisms (Anderson & Denkl, 2010):
Heinrich's ratio is difficult to replicate
too much blame placed on front-line workers (and not, say, management policies)
reducing common/no injury incidents will not necessarily reduce rare/serious incidents
Charles Perrow (1999): Normal Accident Theory
- normal accidents: occur in highly developed technological systems when failure of one component interacts with others, causing a cascading disaster
- caused by three factors:
1. high __________ of interactions (compared to a linear system that behaves predictably)
e.g., the space shuttle was the most complex flying machine ever built, with over 2.5 million parts
2. tight ________: lack of slack or buffer between highly interdependent components
e.g., semiconductor shortage → decrease in vehicle production due to its tightly coupled supply chain
3. ____________ potential
e.g., release of toxic gas that killed and injured thousands of people around Bhopal, India
- criticisms:
too deterministic: if accidents are inevitable, why aren’t there more of them?
cannot predict when or where failure may occur
Davis & Pless (2001):
- editors of the British Medical Journal banned inappropriate use of the word “________”
- most injuries are predictable and preventable, whereas true accidents are unpredictable, and therefore unavoidable
- however, “accident” is still used in the World Health Organization’s International Classification of Diseases (ICD-11): External causes of morbidity or mortality (WHO, 2023)
Michael S. Wogalter (1994):
- to increase safety:
1) remove hazard, 2) guard against hazard, 3) have proper training, 4) post warnings
- critical elements of warnings:
• ______ word (“DANGER”)
• description of ______ (“high voltage lines”)
• description of ____________ (“electrocution risk”)
• ____________ (“do not touch”)
- factors influencing effectiveness of warnings:
|
warning information |
|
|
attention |
|
|
|
_____________ |
|
|
|
beliefs/attitudes |
|
|
|
motivation |
|
|
|
__________ behaviour |
e.g., drinking and driving seen as a threat to personal safety by 81%, but 20% do it (NHTSA, 2010)
Zeitlin (1994): Why don’t people follow safety instructions?
- there may be a breakdown in:
• _____________: user avoids hazard if properly informed
• ________-______: user aware of safeguards, but chooses to ignore them
- why disregard safety information?
• people willing to take (perceived) low-probability risks (expected utility theory)
• people accept more risk if they have actual (or perceived) _______ over the situation
• cost/benefit tradeoff
• __________
• sensation-seeking ___________ trait (Zuckerman, 1994):
- some people seek out thrills and dangerous experiences
- risk-taking behaviours include varied ______ experience, greater use/variety of illicit _____risky driving, risky sports, risky play, etc.
- methods:
• 2 groups: experienced chainsaw users, novices
• half of each group received lectures on aerospace equipment design, half on how safe operating procedures improve industrial safety
• all subjects given 10 chainsaw safety instructions
• given recognition test on safety instructions
• performed tasks with actual ________
- results:
• equipment design vs. industrial safety lectures: no significant effect
• differences between groups:
|
experienced |
novices |
|
|
instruction recognition |
87.5% |
87.5% |
|
instruction compliance |
____% |
70.0% |
- conclusions:
• users __________ instructions
• but experience influenced decision to follow instructions--decreasing compliance
• following instructions seen as unnecessary; overkill
• how to improve safety-compliant behaviour?
Reducing accidents by altering behaviour (Sanders & McCormick, 1993):
• procedural __________
• training
- stress learning of safe behaviours
- ensure learned behaviours transfer to task
- evaluate effectiveness
• feedback
• reinforcement strategies and _________ programs
• increase __________ of risk
Hazard is something that can potentially cause harm to a person or damage property
Risk is the probability that a hazard will cause harm or damage
Risk analysis: estimation of consequences associated with particular errors
- includes estimate of ___________
risk = p (error) × consequences (error)
- can be any sort of risk
e.g., loss of life, money, etc.
- must estimate significance of various consequences
- used to assist many types of decisions:
• estimates of safety
• estimates of probable success
• types of ________ to use to help operators not to miss important errors
- high-probability/low-consequence risks are ______________
The Premortem Strategy (Klein, 1998; 2007):
- a form of _________ risk analysis
- used when organization has made an important decision, but has not yet committed to implementing it
- applies mental ___________: imagining the obstacles that may prevent you from reaching your goal (Oettingen, 1999)
e.g., WOOP: Wish, Outcome, Obstacle, Plan
- and cognitive _________: viewing something from a different perspective, or in a different context
e.g., “Mauritius is a small, insignificant island in the Indian Ocean.” vs. “Mauritius is the largest ocean state in the world.”
- based on ___________ hindsight (Mitchell et al., 1989)
• foresight is predicting the future as an outcome of the current state
• prospective hindsight is is considering a future outcome as if it has already happened, and conceiving of (and generating explanations for) the events that led up to it
• increases accuracy in correctly identifying causes of future outcomes by 30%
- procedure:
Step 0. Know the plan: familiarize decision-making team with premortem plan
Step 1. Preparation: gather all stakeholders in the group together
Step 2. Imagine a ______: leader tells the group: “Imagine that we are a year into the future. We implemented the plan as it now exists. The outcome was a disaster. Please take 5 to 10 minutes to write a brief history of that disaster.”
Step 3. Generate reasons for failure: each person takes 3 minutes to write down reasons why the failure occurred
Step 4. Consolidate the lists: group members take turns reading one of their reasons, which are recorded
Step 5. Revisit the plan: address items of greatest concern, looking for ways to strengthen the plan
Step 6. Periodically review the list: to deal with emerging problems
- example:
• simulated army helicopter mission: cross battle lines into enemy territory, drop off troops, return to base
• to avoid artillery, mission had a 1-minute window to deliver troops at drop zone--very difficult due to terrain, anti-aircraft batteries, etc.
• only 1 crew out of 20 was successful
• premortem established ___________ plans: what to do if helicopter arrived too early or too late
- pros & cons:
evidence:
▸ reduces overconfidence in desired outcomes (Veinott et al., 2010)
▸ meta-analysis shows mental contrasting increases goal attainment (Wang et al., 2021)
may overcome __________?
identifies potential problems
makes all team members feel valued
lacking in direct experimental evidence
Peltzman effect (1975):
- hypothesized tendency for people to react to a driving safety regulation by increasing other risky behaviour, __________ some or all of the benefit of the regulation
- safety devices in automobiles have saved occupants’ lives, but at the expense of more pedestrian deaths and more nonfatal collisions
- this is a form of “behavioural adaptation” or “risk compensation” (OECD, 1990)
Gerald Wilde (1994, 2001, 2014):
- the degree of risk-taking behaviour, and the magnitude of loss due to accidents and lifestyle-dependent disease are __________ over time--unless target risk level changes
- ______ ____: level of risk a person accepts to maximize overall expected benefit from an activity
- determined by four categories of motivating (i.e., subjective utility) factors:
1. expected benefits of comparatively _____ behaviour
2. expected costs of comparatively _____ behaviour
3. expected benefits of comparatively ____ behaviour
4. expected costs of comparatively ____ behaviour
- greater risk as #1 and #4 increase; and/or as #2 and #3 decrease
Evidence:
• the introduction of “child-proof” medicine bottles in 1972 resulted in poisonings in children under age 5 increasing by 3,500 per year
• laws banning cell phone use while driving have not reduced crashes
• people drive closer to bicyclists who are wearing _______
• motorcyclists drive faster when wearing full leathers, compared to wearing only __________
• stricter NASCAR regulations have led to an increase in crashes
• hockey players got more penalty time after switching to helmets with visors
• when Sweden switched from driving on the left side of the road to driving on the right side; traffic fatalities decreased by 17% in the next 12 months (but after 2 years, fatalities increased to their original levels)
• when speed limits were increased and fines for speeding reduced in a driving simulator, driving speed increased but collision frequency remained constant
Adams (1999, 2001): risk compensation in seat belt use
- seat belts reduce chances of death by 41%
- does ___________ reduce risk?
- laws passed in over 80 jurisdictions worldwide
- only measurable change in the UK:
• use rose from 40% to 90% 
fatalities decreased 20%
• confounding factor: firm campaign against drunk driving introduced at the same time
• fatalities dropped significantly between 10 p.m. and 4 a.m. (“_____ driving hours”)
• UK drop due to decrease in drunk drivers, not seat belt laws
- specific changes in the UK:
• deaths of front seat belt-wearers decreased by 200
• but pedestrian, cyclist, rear-seat passenger deaths _________ by 800
- risky driving presumably increased to __________ for the safety of the seat belts
Aschenbrenner & Biehl (1994): The Munich Taxicab Experiment
- anti-lock braking system (ABS) prevents wheels from locking up under extreme braking conditions
- advantages: improved steering control during rapid deceleration, especially on slippery road surfaces
Part I
- part of taxi fleet equipped with ABS
- otherwise, all vehicles were identical and were driven in the same weather/traffic conditions; drivers were randomly assigned to vehicles
- drivers were informed whether the vehicle had ABS or not
- RHT predicted ___________ __________ in drivers such that they would maintain a constant likelihood of collisions per hour of driving
- 747 collisions investigated
- no statistically significant difference in collisions
Part II
- researchers installed accelerometers (measure g-force of acceleration/deceleration)
- more rapid deceleration in ABS cars (but more rapid acceleration in control cars)
Part III
- driving style observed; evaluated on rating scales
- 113 trips made (57 in ABS, 56 in non-ABS)
- observers did not know whether car had ABS or not
- drivers in ABS cars:
• made sharper turns in curves
• were less ________ in lane-holding behaviour
• proceeded at a shorter forward sight distance
• made more poorly adjusted merging manoeuvres
• created more “_______ _________”: one or more traffic participants must take swift action to avoid collision
• ABS cabs driven ______ at one of 4 measuring points along the route, compared with non-ABS cabs
Followup (Wilde, 1994)
- next year, ABS cars had more collisions under slippery conditions
- the following year, all collisions _________
- cab company made drivers ___ for part of the repairs (increased expected cost of risky behaviour)
Criticisms of RHT:
criticized as not falsifiable (Hoyes & Glendon, 1993; Trimpop, 1996)
contrary evidence:
• no relationship found between helmet use and risky motorcycle riding (Oullet, 2011)
• speed limit laws have decreased fatalities per unit time of driving (Evans, 1986)
people do not have enough knowledge, ability, or attention to adjust their behaviour to maintain a constant level of risk (Robertson & Pless, 2002)
Leonard Evans (1999): reviewed data on ABS
- overall, ABS is safer:
• 10% reduction in crashes on ___ roads
• 22% reduction in pedestrian crashes
• 32% reduction in rear-enders
- but drivers with ABS:
• 30% increase in being rear-ended
• drive in weather conditions previously _______
• drive ______ than previously
• follow more closely
• driver behaviour may also increase rollover risk
- similar data on effectiveness of brakes as far back as ____!
- traditional approach uses triple-E traffic calming: engineering, enforcement, education (e.g., speed bumps, photo radar, public service announcements)
- traffic engineers typically _________ vehicle traffic from pedestrian/bicycle traffic
- with increasing traffic, typically more roadways are devoted to vehicles
(ironically, adding more roads or lanes can decrease overall traffic flow due to induced demand; this is Braess’s paradox, 1968)
- regulations are indicated by traffic signs (about 70% of which are _______ by drivers)
- unintended consequences:
• moving people off the streets made them ____ safe when they had to cross the street
• human scale of the urban environment is obliterated
- counterintuitive solution: Shared space (Hans Monderman, Fryslân Province, The Netherlands, 2005), a.k.a. complete streets or woonerf
• has more ___________ of vehicle and pedestrian traffic
• intersections have no traffic control devices: signals, signs, road markings, or pedestrian crossings
• the architecture of the road determines traffic flow
• the resulting ambiguity forces drivers to be more cautious and pay more attention: “______ is safer”
Evidence:
Tom Vanderbilt (2008): Traffic: Why We Drive the Way We Do
- counterintuitive advantages of traffic circles and roundabouts:
• _____ __ traffic (on average), because there is no stopping for red lights
• vehicles, when moving, travel more slowly
• risk of ______ collisions is virtually eliminated
- roads with a 55 mph limit can handle ____ traffic than those with a 70 mph limit: you have to leave more space the faster you go, so the roads are used less efficiently
Shared space is being adopted in Austria, Denmark, France, Germany, Spain, Sweden, the UK, and the USA:
• Drachten, Netherlands: main intersection handles 22,000 vehicles/day; average of 9 casualties/year dropped to 1/year
• Christianfield, Denmark: serious/fatal collisions dropped from 3/year to ____
• Norrköping, Sweden: no collisions, mean traffic speed slowed by 5 km/h, liveability increased
• Wiltshire, UK: removing centre line caused 35% decrease in collisions
• Brighton, UK: less vehicle traffic (93% decrease), slower vehicle traffic (16 km/h less), increased cycle and pedestrian traffic (93% and 162%)
• West Palm Beach, Florida: slower traffic, fewer collisions, _______ trip times
Practical lessons learned from Shared Space (2005):
1. the road tells the story: the road and its surroundings indicate which _________ is appropriate and required
2. make room for ______: encourage interaction, and facilitate eye contact
3. people have a say: incorporate citizen input; have them cooperate with government
4. details can make or break the design: choose materials that suit the buildings and landscape
5. better chaotic than pseudo-safe: do not try to remove that ______ feeling, but use it to best effect
Pros and Cons:
better traffic flow: removing traffic lights allows freer traffic flow, which leads to shorter travel times and less pollution
reduced __________: drivers approach intersections more slowly; concentrate more on their surroundings
__________: less visual clutter from signs, traffic lights, etc.
e.g., Drachten installed decorative fountains in former intersections
less “________”: less interference from rules/regulations; drivers make their own decisions
vital information/signs not available to drivers
lack of railings, curbs and barriers is hazardous to _____ pedestrians
e.g., assistance dogs are trained to stop at curbs
design encourages some drivers to “bully” cyclists and pedestrians, who experience feelings of anxiety/fear
effects are due to _______; once shared space becomes the norm, drivers will be less careful