Based on MIT CSAIL Social Value Orientation research, MIT AgeLab driver behaviour studies,
Dual Process Theory, risk homeostasis, young and older driver neuroscience, and behaviour change science
Nobel laureate Daniel Kahneman's Dual Process Theory is fundamental to understanding driver behaviour. Kahneman identified two modes of thinking: fast, automatic, intuitive System 1, and slow, deliberate, analytical System 2. The critical insight for road safety is that driving — especially for experienced drivers — is almost entirely System 1, which means it is fast and efficient but also prone to systematic biases and errors that System 2 would catch.
Benefits: Fast enough to respond to hazards. Frees conscious mind for route planning and conversation.
Risks: Over-familiar patterns, confirmation bias, optimism bias — all System 1 phenomena.
When it activates: New situation, something seems wrong, learning mode, high-stakes decision.
Risk when absent: Familiar route with changed conditions (roadworks, new hazard) — System 1 uses the old "safe" template.
Advanced driving training (IAM, ROSPA, Garda pursuit training) works by deliberately engaging System 2 on roads where System 1 has become too comfortable. IPSGA and commentary driving are System 2 techniques that make unconscious decisions conscious again — recalibrating the automatic responses that have drifted toward complacency.
Crashes are disproportionately likely on roads drivers know well — their daily commute, routes driven hundreds of times. System 1 has learned these routes as "safe" and stops scanning for new hazards. The child who steps off a pavement for the first time on a familiar road is invisible to a driver in full System 1 mode.
Humans are systematically poor at estimating risk. We over-fear dramatic, vivid, unfamiliar dangers (plane crashes, shark attacks) and under-fear familiar, everyday ones (car journeys, unhealthy diet). This is not stupidity — it is how the brain's threat-detection system evolved. For road safety, it means drivers routinely underestimate the risks of their own habitual behaviours.
The universal human tendency to believe we are less likely than average to experience negative events. Studies consistently show that when asked "What is your risk of a road crash compared to average?", over 70% of drivers rate themselves as below-average risk — a statistical impossibility.
For road safety this means: most drivers believe the crash statistics apply to "other drivers" — not to themselves. This directly reduces motivation to change behaviour.
The more often a driver does something without incident, the less risky it feels. Driving the same route at 100 km/h every day for 5 years with no crash → that speed on that road feels "safe." The actual physics of a crash has not changed; the perception has.
Risks we "dread" feel bigger than their statistics suggest. Risks we are exposed to daily feel smaller. Flying feels riskier than driving — even though per km, driving is 20× more deadly — because plane crashes are vivid and rare, car crashes are invisible and common.
MIT AgeLab research on hazard perception training found that giving drivers accurate risk statistics — "at your speed, if a child stepped out 20 metres ahead, you would hit them" — measurably improved risk calibration and subsequent behaviour. Abstract statistics ("crashes cause X deaths per year") did not change behaviour. Personal, specific, concrete risk information did.
Drivers consistently overestimate their ability to control outcome in risky situations — "I'm a good driver, I could handle it." Control illusion inflates confidence and reduces precautionary behaviour, particularly in high-skill domains where drivers have seen themselves perform well in normal conditions.
Prof. Gerald Wilde (Queen's University) proposed Risk Homeostasis Theory in 1994: drivers do not try to minimise risk — they seek a subjectively acceptable target level of risk. When safety improves (better brakes, safer roads), drivers unconsciously increase their risk-taking to maintain their preferred level. This theory is debated, but the core phenomenon — behavioural adaptation to safety measures — is well-documented and important for safety design.
Think of a thermostat: the system continually adjusts to maintain a target temperature. Wilde's model treats risk perception as a thermostat. Safety improvements → driver feels "safer" → driver speeds up, follows closer, accepts more gaps — until perceived risk returns to the target level.
The implication: engineering solutions alone may not reduce crash rates proportionally, because driver behaviour compensates. The total risk in the system may remain roughly constant.
When ABS (anti-lock braking) was introduced, it was expected to reduce crash rates significantly. Wilde predicted drivers would follow more closely on wet roads because braking felt "safer." Studies confirmed shorter following distances in ABS-equipped vehicles, partially offsetting the safety gain.
Motorways are statistically the safest roads per km. Yet some drivers drive faster on motorways than rural roads, partly because the road environment (wide lanes, barriers, no junctions) makes high speed feel safe. The safety feature creates the speeding behaviour.
Supporting evidence: Cycle helmets → riskier cycling behaviour. Seatbelts → some evidence of faster driving. ABS → shorter following distances. Speed cameras removed → speeds increase immediately.
Counter-evidence: Many safety improvements DO reduce absolute crash rates (AEB, median barriers). Not all drivers show behavioural adaptation. The effect is real but does not fully offset safety gains in most cases.
Safety measures work best when they reduce crash severity rather than rely on changing driver behaviour. A median barrier prevents head-on crashes regardless of driver behaviour — no adaptation possible. AEB brakes for you — no adaptation. These are more robust than measures relying on drivers to maintain changed behaviour.
This is why Safe System engineering prioritises physical protection over driver behaviour change alone.
MIT CSAIL (Computer Science and Artificial Intelligence Lab) applied Social Value Orientation theory to driving to explain why drivers respond so differently to the same traffic situations. SVO is a personality dimension that measures whether a person prioritises their own outcomes only (individualist/competitive) or also values others' outcomes (prosocial/cooperative). In traffic, this predicts whether a driver will let others merge, maintain safe gaps, or compete aggressively for position.
Prosocial (≈60% of drivers): Seeks good outcomes for self and others. Yields, maintains gaps, cooperative in merging. Values social harmony in traffic. Traffic flows better when these drivers interact.
Individualist (≈30% of drivers): Focuses on maximising own journey time with little concern for effect on others. Will merge aggressively, exploit gaps, but not necessarily maliciously — just self-focused.
Competitive (≈10% of drivers): Seeks to outperform other drivers — their measure of success is relative position, not absolute journey time. Will tailgate, refuse to yield, cut up. These drivers actively worsen traffic flow for everyone.
MIT CSAIL researchers used instrumented vehicles and the CarTel mobile sensing system to measure driving behaviour in Boston traffic. They correlated driving behaviour patterns with SVO scores measured in pre-drive psychological testing. Finding: competitive SVO drivers were 40% more likely to make aggressive lane changes, 60% more likely to drive within 1 second following distance, and had 2.3× the near-miss rate of prosocial drivers — despite reporting equal subjective confidence in their driving ability.
MIT CSAIL modelling showed that if 15% of drivers exhibit competitive SVO in a dense traffic stream, it degrades overall flow for ALL drivers by up to 30%. A small minority of competitive drivers imposes significant time and risk costs on the majority. This is the "tragedy of the commons" in traffic.
This finding supports the case for automated vehicles — an AV with prosocial programming could individually outperform human drivers and collectively improve system efficiency.
SVO in traffic can shift situationally: drivers under time pressure shift toward individualist/competitive behaviour. Drivers who have recently been cut up show competitive retaliation patterns. Running late is a temporary SVO shift that permanently increases crash risk. The safest drivers maintain prosocial behaviour even when provoked or rushed.
Aggressive driving sits at the intersection of competitive SVO, situational frustration, stress, and disinhibition. It ranges from persistent tailgating and aggressive lane changes (aggressive driving) to verbal confrontation, vehicle ramming, and physical assault (road rage). Research shows the transition from frustration to aggression follows a predictable psychological pathway — and understanding it helps drivers recognise and break the cycle.
Another driver cuts in, moves too slowly, fails to signal. The event is interpreted as intentional rather than accidental ("that idiot deliberately did that").
We blame character, not situation. The slow driver is "inconsiderate" — we don't consider they might be lost, ill, or learner. This is the Fundamental Attribution Error.
Cortisol and adrenaline rise. Heart rate increases. The amygdala (emotional processing) activates. Prefrontal cortex (rational decision-making) is suppressed. The driver is physiologically primed for fight-or-flight.
Anonymity and physical protection of the car create disinhibition. Behaviours that would be unthinkable face-to-face (screaming abuse, physical intimidation) become thinkable behind glass.
Tailgating, flashing lights, cutting up, verbal abuse, or physical confrontation. The focus shifts entirely from safe journey completion to "winning" the interaction. Both parties are now at dramatically elevated crash risk.
Anger and emotional arousal produce measurable driving impairment comparable to mild alcohol consumption. Studies show:
• Reaction time increases by 20–40%
• Following distance decreases
• Risk-taking decisions increase
• Attention narrows (tunnel vision on the target vehicle)
• Other hazards are missed
An angry driver is an impaired driver. The impairment is invisible to anyone watching from outside.
Young driver crash rates are not simply the result of inexperience — they have a neurological basis. The adolescent and young adult brain is structurally different from the mature adult brain in ways that directly affect risk assessment, impulse control, and sensitivity to social influence. Understanding this biology transforms the conversation from "irresponsible young people" to "predictable brain development mismatch."
The PFC is the brain region responsible for: impulse control, risk assessment, consequence evaluation, long-term planning, and overriding emotional reactions. It is the brain's rational executive.
Critical fact: the PFC is not fully mature until approximately age 25. It is the last brain region to complete development.
This means a 17-year-old driver has an incomplete neural tool for risk assessment — regardless of how mature or intelligent they are.
While the PFC matures slowly, the limbic system (emotion, reward-seeking) is fully developed in adolescence — and is actually hyperactivated during teenage years. The result: strong reward-seeking and emotional reactivity with inadequate inhibitory control. A sports car, peer dares, or emotional excitement can override safety reasoning entirely.
Adolescent and young adult brains show heightened activation of the dopamine reward system in social contexts. The peer approval reward is neurologically more powerful than in adults. This explains why young drivers take risks in front of peers that they do not take alone — the social reward literally overrides risk inhibition at a neurological level.
The peer passenger effect: Each additional teenage peer passenger multiplies a novice driver's fatal crash risk by approximately 1.5×. Three passengers → 3.4× baseline crash risk.
MIT AgeLab's longitudinal study followed novice drivers from licence acquisition. Key finding: the first 6 months of independent driving carries 5× the crash rate of experienced drivers. The crash rate halved after 12 months and continued declining to age 25. The curve follows PFC maturation. This provides the scientific basis for Graduated Driver Licensing restrictions during the high-risk developmental window.
MIT AgeLab is internationally recognised as a world leader in older driver research. Their LNTP (Life-Span Naturalistic Driving Programme) and Advanced Vehicle Technology studies provide the most comprehensive picture available of how driving capability changes with age — and crucially, how older drivers successfully adapt. The picture is more nuanced than "older drivers are dangerous."
MIT AgeLab's longitudinal naturalistic data shows that older drivers who engage in deliberate self-regulation — voluntarily avoiding their known challenging conditions — can maintain safe driving well into their 80s in many cases. The most dangerous older drivers are not those who drive slowly, but those who lack insight into their own limitations and continue driving in conditions that exceed their current capability.
Driving cessation for older adults is associated with increased social isolation, depression, and even reduced life expectancy. The goal should not be to remove older drivers from roads — it is to help them drive safely for as long as possible. Assessment, training, vehicle adaptations (wider mirrors, better lighting, adaptive cruise) are more humane and effective than blanket restrictions.
Popular culture recognises "the aggressive driver," "the careful driver," and "the reckless driver" as personality types. But does the scientific evidence support stable driving personalities? MIT CSAIL and AgeLab research — combined with Big Five personality psychology — shows that personality does predict driving behaviour, but the relationship is complex and context-dependent.
| Trait | High score effect |
|---|---|
| Neuroticism | Anxious driving, over-caution, poor emergency response |
| Extraversion | Higher speeds, more social driving, more distractions accepted |
| Openness | Little direct effect on crash risk |
| Agreeableness | More cooperative, lower aggressive driving |
| Conscientiousness | Strongest predictor of safe driving — rule-following, attention to detail |
Sensation-seeking — the desire for novel, intense experiences — is strongly associated with speeding, running red lights, and drink driving. Highest in 18–24-year-old males. Declines with age. Partially genetic. Correlates with competitive SVO in traffic.
Sensation-seeking drivers do not respond well to fear-based campaigns — the arousal of fear can paradoxically increase the appeal of the forbidden behaviour. Educational approaches emphasising skill and mastery work better.
MIT CSAIL researchers used machine learning on naturalistic driving data to identify natural clusters of driving behaviour. They found three robust styles:
Calm/Safe: Smooth acceleration, long following distances, low speed variance. Low crash rate.
Normal: Average profile, situationally variable. Average crash rate.
Aggressive: Hard braking, rapid acceleration, short following distance, high speed variance. 3× crash rate of calm drivers.
Critically: style was stable across time and road types. It is a genuine behavioural trait, not just situational.
Personality is a tendency, not a destiny. High sensation-seekers with good training and strong safety culture show significantly lower crash rates than untrained sensation-seekers. Personality identifies who needs more targeted intervention — it doesn't make intervention futile.
Drivers do not step into a vehicle as emotionally neutral operators. They bring their day's stress, their anxieties, their grief, their anger, and their elation. Research shows that emotional state at the time of driving significantly affects risk-taking, attention, and decision quality — and that the vehicle itself can amplify emotional states through isolation, perceived anonymity, and the frustrations of traffic.
Anger: Increases speeding, tailgating, aggressive manoeuvres. Narrows attention. Elevated cortisol impairs judgment. The most dangerous single emotional state for driving.
Anxiety/Worry: Rumination occupies working memory. Driver may be technically competent but cognitively elsewhere. Slow, hesitant responses at junctions. Paradoxically, very anxious drivers are also crash-prone.
Sadness/Grief: Impairs attention, slows processing, reduces motivation to drive carefully ("nothing matters"). Grief is documented in driving crash data — bereaved drivers have elevated crash risk in the 3–6 months after loss.
Elation: Euphoria and high positive affect can also impair driving — through overconfidence, risk underestimation, and reduced attention to mundane driving tasks.
Excitement: Similar to anger in physiological terms — elevated heart rate, reduced inhibition. Young drivers receiving good news, or excited by music, show similar impairments to mildly angry drivers.
MIT AgeLab equipped research vehicles with physiological monitoring (heart rate, skin conductance, facial expression cameras) alongside driving performance metrics. They found that emotional state at journey start predicted driving behaviour throughout the journey. Drivers who were emotionally dysregulated at journey start showed elevated hard braking events, lane departures, and speed variance compared to emotionally neutral baseline drives — even 30+ minutes into a journey.
If you have just had a major argument, received shocking news, or are in a state of emotional agitation — wait 5 minutes before driving. Take deep breaths. The journey can wait. Your emotional state at the wheel is a safety-critical variable.
Habit formation is one of the most powerful forces in human behaviour. When a behaviour is repeated in the same context, it transitions from deliberate (System 2) to automatic (System 1). This is how we learn to drive — and it is how driving becomes efficient. But habits are extraordinarily resistant to change, and they continue running even when context has changed in a way that makes the old habit dangerous.
Cue → Routine → Reward. Every habitual driving behaviour follows this structure. The morning commute cue triggers the routine of "drive the familiar route at familiar speed." The reward is efficient, low-effort journey completion.
Habit neurological pathways are in the basal ganglia — below the cortex. Once established, they can be suppressed by cortical intention but not erased. Under stress or cognitive load, cortical suppression fails and the old habit reasserts itself. This is why drivers "revert" to unsafe behaviours when tired or distracted.
A driver habitually rolls through the stop sign on their quiet street at 7am. One day a cyclist appears for the first time. The habit continues — the cyclist is hit. The stop sign had been rolled through 1,000 times safely. The habit was built on the absence of hazard, not on safety of the manoeuvre itself.
MIT research on commuting behaviour found that driving habits are among the most stable human behaviours. Even when commuters change jobs (and therefore route), they revert to old route patterns within weeks if there is any possibility of using the original route. The habit cue (leaving the house in the morning) is so strong it overrides new intentions. This has significant implications for introducing new road layouts — drivers will use the old routing behaviour for months before adapting.
Habit formation can be harnessed for safety as well as against it. Consistently performing a safety routine (mirror check before moving off, 3-second commentary at junctions) in the same context eventually automates that behaviour — it becomes a default, low-effort habit rather than requiring conscious effort each time.
This is why the best driving training emphasises consistent routines rather than just knowledge of rules. Rules fade; habits persist.
The decision to drive while impaired is not usually made by a person who thinks "I'll drive drunk." It is made by a person whose judgment — including their judgment of their own impairment — has been degraded by the substance they have consumed. Understanding this self-reinforcing impairment helps explain why awareness alone does not prevent drink driving, and what interventions actually work.
| BAC | Driving Effects |
|---|---|
| 0.02% | Divided attention begins to degrade, relaxation reduces vigilance |
| 0.05% | Reaction time +15%, hazard detection reduced, risk tolerance increases |
| 0.08% | Crash risk 4×, coordination and brake response significantly impaired |
| 0.10% | Crash risk 7×, severe judgment impairment |
| 0.15% | Crash risk 25×, major motor and visual impairment |
Alcohol simultaneously impairs driving ability AND reduces the driver's perception of their own impairment. A drunk driver feels more confident, not less — lower anxiety, reduced self-monitoring, overestimation of competence. The person least qualified to judge whether they should drive is the intoxicated person deciding whether to drive.
Alcohol metabolises at approximately 1 unit per hour. After heavy drinking (10+ units), BAC can still be above 0.05% the following morning. Many drink-drive prosecutions occur the morning after a night out — the driver genuinely believes they are "fine."
Sleep does not speed metabolism. The only thing that removes alcohol is time and liver function. Coffee, water, and food have no effect on BAC.
THC (cannabis) increases crash risk by approximately 2×. Unlike alcohol, cannabis impairment cannot be precisely measured by any roadside test currently in use. Critically, cannabis users significantly underestimate their impairment — studies show users rate their driving as normal when simulator performance shows significant degradation. Cannabis impairs: reaction time, lane keeping, speed regulation, and risk assessment. Combined with alcohol, the effect is multiplicative, not additive.
Legal consequences deter drink driving when perceived detection probability is high. In Ireland, mandatory alcohol testing checkpoints (MATs) are the single most effective deterrence measure because they create visible, credible detection risk. When enforcement is invisible, deterrence fails regardless of penalty severity.
Behaviour change is one of the most studied topics in psychology, and the road safety field has repeatedly failed to apply its findings. Decades of shock advertising, speed awareness courses, and knowledge campaigns have produced limited results. Understanding what actually changes behaviour — versus what feels like it should — is essential for effective road safety work.
MIT AgeLab research on behaviour change in older drivers found that "nudges" — small environmental design changes that make safe choices easier — were more effective than traditional education. Examples: making the seatbelt buckle more visible and accessible (increased use by 8%), changing default navigation announcement to include road condition alerts (increased compliance with speed adjustments). Design for safety, not just education for safety.
The Goals for Driver Education (GDE) matrix is the EU's evidence-based framework for what comprehensive driver education must achieve. Developed by Hatakka et al. and adopted across Europe, it identifies four levels of driving competence — from basic vehicle control to self-awareness and lifestyle. Most driving tests only assess the bottom two levels; crashes are mostly caused by failures at the top two levels.
Level 1 — Vehicle manoeuvring: Physical control of the vehicle. Steering, braking, gears, parking. This is what driving tests primarily assess.
Level 2 — Mastering traffic situations: Handling interactions with other road users. Junctions, overtaking, merging, following distance. Most of road safety education operates here.
Level 3 — Goals and context of driving: Why am I driving? What goals am I trying to achieve? Am I in the right state? Should I even be driving now? Trip planning and journey decisions.
Level 4 — Self-awareness: What are my personal strengths and weaknesses as a driver? How do my personality, emotions, attitudes, and values affect my driving? This is the level most missed by traditional education.
Research consistently shows that most crashes do not result from lack of vehicle control skill (Level 1) or even traffic knowledge (Level 2). They result from Level 3 and 4 failures: driving when tired, rushing because running late, driving while emotionally aroused, overestimating one's own skill, normalising risky behaviour.
A driver who knows they are a sensation-seeker and has high aggression (Level 4 self-awareness) and therefore consciously applies more conservative strategies is safer than a technically excellent driver who is blind to their own emotional vulnerabilities.
Ireland's Approved Driving Instructor curriculum and the Essential Driver Training (EDT) programme both incorporate GDE Level 3 and 4 elements: journey planning, attitude assessment, understanding own risk factors. These are tested in the driving test via discussion and reflective questions — not just vehicle handling.
Traffic is a profoundly social environment — we are continuously influenced by the behaviour of drivers around us, by cultural norms about "normal" driving, and by the expectations of people inside our vehicle. MIT CSAIL research on the social dynamics of traffic shows that individual driver behaviour cannot be fully understood without understanding the social context in which it occurs.
Drivers calibrate their speed partly to the speeds of surrounding vehicles. If everyone around you is travelling at 130 km/h on a 120 km/h motorway, travelling at 120 km/h feels actively dangerous (risk of rear-end collision). Social norm has overridden the legal limit.
This creates a self-reinforcing cycle: if most drivers speed, new drivers learn that speeding is the "real" norm. Compliance with limits becomes the deviant behaviour.
The most common justification for unsafe behaviour is descriptive norm reference: "everyone runs that junction," "everyone speeds on that road," "everyone uses their phone." Research shows perceived peer behaviour is more predictive of individual behaviour than personal attitudes or legal knowledge.
Interventions that correct over-estimated norms ("actually, 70% of drivers on this road comply with the speed limit") are effective specifically because they undermine this rationalisation.
MIT CSAIL's CarTel project (mobile traffic sensing) ran an experiment where drivers received feedback showing how their driving compared to other drivers on the same roads. Drivers who received feedback that they were in the top 20% most aggressive (measured by hard braking, acceleration, speed) showed a 25% reduction in aggressive driving events over 4 weeks. Social comparison — particularly downward comparison ("you are more aggressive than most") — is a powerful motivator for change.
Parents are the single most influential model for their children's future driving behaviour. Research shows parental driving style is more predictive of child driving style than formal instruction or peer influence. The "do as I say, not as I do" parent creates a dangerous mismatch between the rules they teach and the norms they demonstrate.
We covered the neuroscience of fatigue in Module 01. This slide focuses on the psychology of the decision to drive fatigued — why people knowingly get behind the wheel when tired, and what that decision process looks like. It is rarely straightforward recklessness — it is usually a series of small rationalisations, each of which seems reasonable, leading to a collectively dangerous outcome.
5–6 hours sleep per night over a week creates massive sleep debt. Driver feels "a bit tired" but not "dangerously tired." The gap between felt tiredness and actual impairment is large.
Distance rationalisation. Microsleep can occur within 10 minutes of driving onset. Distance does not protect — sudden onset at any point on the journey is equally possible.
Normalisation of deviance. Previous survival creates false safety signal. Prior exposures to the risk with no consequence increase confidence, not safety.
Countermeasure rationalisation. These measures provide minutes of relief — not sufficient for a journey of any length. Driver overestimates their effectiveness.
"I have to get home." "There's nowhere to stop." "I can't afford a hotel." Economic and logistical pressures override safety judgment — especially at night on motorways.
After 20 hours awake, driving performance is equivalent to a BAC of 0.08% — the legal drink-drive limit. After 24 hours, performance matches BAC 0.10%. Unlike alcohol, severe fatigue does not produce the subjective experience of being impaired. Fatigued drivers feel they are driving normally up to the point of microsleep. The impairment is invisible from the inside.
Insurance telematics (black boxes), fleet monitoring systems, and smartphone driving apps represent the most significant recent development in driver behaviour change. By providing drivers with personalised, accurate, immediate feedback on their own driving behaviour, telematics addresses the single biggest barrier to behaviour change: the gap between what drivers believe they do and what they actually do.
Telematics is effective because it addresses the psychological barriers to change:
• Optimism bias: "I'm a good driver" is confronted by personal data showing actual speeding events.
• Social comparison: Showing your score relative to peers activates competitive prosocial motivation.
• Immediacy: In-car real-time alerts provide consequence in the moment — not after a crash.
• Gamification: Score improvement creates a reward loop that sustains motivation beyond initial awareness.
Telematics measures inputs (hard braking, speed) not outcomes. A driver can have a perfect telematics score while driving inattentively at legal speeds. Cognitive distraction, fatigue, and inattention are not captured. Telematics improves the measurable behaviours it measures — it doesn't make a driver attentive or alert.
The value of this module is not in describing other drivers' psychology — it is in recognising your own psychological vulnerabilities and designing your behaviour around them. Every driver has biases, habits, and emotional triggers. The difference between safe and unsafe drivers is often not the presence or absence of these vulnerabilities — it is the level of self-awareness about them.
Research on advanced drivers (IAM, ROSPA) shows they share specific psychological traits: high self-awareness about their own limitations, consistent prosocial SVO in traffic, strong commitment to journey completion over speed, and deliberate use of System 2 thinking in familiar environments. These traits are trainable — they are not personality fixed. The expert driver's key skill is knowing when their automatic thinking needs to be overridden.
| Researcher / Model | Contribution | Key Concept |
|---|---|---|
| Daniel Kahneman | Dual Process Theory (2011) | System 1 vs. System 2 thinking in driving decisions |
| Gerald Wilde | Risk Homeostasis Theory (1994) | Drivers seek a target risk level; safety measures can be offset by behaviour change |
| MIT CSAIL | Social Value Orientation in traffic | Prosocial vs. competitive drivers; social dynamics of traffic flow |
| Bryan Reimer (MIT AgeLab) | Distraction & automation research | 27-second cognitive residue; PERCLOS; ADAS complacency |
| Joseph Coughlin (MIT AgeLab) | Ageing and mobility research | Older driver self-regulation; LNTP longitudinal data |
| Hatakka et al. | GDE Matrix (2002) | 4-level driver education from vehicle control to self-awareness |
| Paul Slovic | Risk perception (1987) | Systematic biases in how humans estimate risk |
| Laurence Steinberg | Adolescent risk neuroscience | PFC immaturity, social reward sensitivity in young drivers |
| Michie et al. | Behaviour Change Wheel (2011) | COM-B model: capability, opportunity, motivation needed for change |
Crashes happen not just because of what drivers do — but because of who they are, how they think, and what they feel at the moment they drive. Psychology is not a soft topic in road safety — it is the central one.
End with the GDE Level 4 self-assessment questions from Slide 18. Ask participants to privately identify their single biggest psychological vulnerability as a driver. Then ask: "What one structural change could you make to your driving environment that would make that vulnerability less dangerous?" This is the practical payoff from the entire module.