Traffic Psychology

Developing a Driver Behaviour questionnaire

Distinction between skill and style of driving

Proximal

Impact of stimuli can start at time of action

Potential influence of impulsivity

Offenders produced more commission errors than non-offending age matched controls More sensitive to group differences than psychometric measures Respond Respond Inhibit Response Inhibit response

Task Capability Interface Important concepts: Loss of control

how demanding it is to complete driving related behaviour e.g. steering car round a corner

Developing a Driver Behaviour questionnaire Distinction between skill and style of driving Drivers skills/performance

include information-processing and motor skills; both improve with practice and training

Task Capability Interface Model

Using models to explain driver behaviour 1. perception 2. expectation 3. judgement 4. memory 5. decision/planning 6. car handling/driver action 7. errors, assesment, and conclusion 8. rating of complexity 9. assesment of risk

Influencing ATD and RT

When late, immediate goals will change as will the effort the driver imparts Jamson (2008) drivers drive closer to driver in front when emotions are elevated; Gormley, Walsh & Fuller (2008) under blue lights ambulances and fire trucks crash rates increase 3 fold

Exceeding threshold TDi

Current TDe can be increased as the driving task becomes more complicated in an expected fashion e.g. route requires navigation of a complex junction TDe increased due the addition of an unexpected additional task e.g. having to brake due to car in front suddenly braking DC becomes impaired e.g. due to alcohol consumption

Biological- Adolescent rewards system

Encourages sensation seeking or risk taking Number of dopamine receptors peeks in areas such as NAcc and VS Adolescence is a period of increased reward associated with novel/exciting experiences Sensation seeking correlated with wide range of risky behaviours: Risky driving, taking drugs, unsafe sex, gambling, delinquency

Skill/performance improves reasonable incrementally

Hanfling skills developing quite quickly Hazard perception skills takes a bit longer Higher order skills arguably takes longest of all skills styles/behaviour becomes established over years but does not necessarily ger safer with experience with less concern for safety and habitually driving with narrower safety margins

Types of collisions older drivers

If collision type changes by age category indicates something unique happening for that group Across most age groups majority of car driver KSIs occur away from junctions but amongst the over 70s there is an even split Drivers aged over 80 were relatively more likely to be involved in KSIs when they were turning right (across traffic)

Transtheoretical Model of Change

Individuals go through stages which may subsequently contribute to behavioural change - the stage you are in will impact the likelihood of behavioural change (Prochaska, Norcross & DiClemente, 1994) - Pre-contemplation (unaware of the need or benefit to change) - Contemplation - Preparation - Action - Maintenance

Origins of the problem: young drivers

Inexperience Biological Neurological development: Reward vs control Physiological maturity of the body Cognitive How they view risk Positivity bias Social Influence of peers inside and outside the car

Reliability of the DBQ

Internal reliability measured using an reliability coefficient such as Cronbach's alpha For DBQ (Özkan, Lajunen, & Summala, 2006): Errors (.84), violaTIons (.83) but Mistakes (.81), lapses (.67) Ordinary violaTIons (.79), aggressive violaTIons (.74) Reflects that fact that item number impacts on reliability with reliability increasing as item number increases

The problem of calibration

Overestimation of safety margin leads to e.g., excessive speed, short following distances, and the performance of risky manoeuvres with inevitable consequences (Kuiken & Twisk, 2001)

Lewis-Evans and Charlton (2006)

Participants drove faster on wider roads Reported the roads to differ in terms of the amount of traffic, number and severity of curves, and amount of signage and road delineation - but these were the same across all 3 road types Likely that implicit perception of risk and possibly perceived TDi contributed to speed choice

Multiphase Driver Education programmes

• Mix learning in driving school with learning during independent driving after licensure - Will involve two/three phases e.g. Norway and Finland - Post licensure education will focus on driving in difficult situations such as at night but also concentrates on selfassessment and hazard perception

calibration

Young drivers particularly overestimate their capability (De Craen, Twisk, Hagenzieker, Elffers, & Brookhuis, 2011) Young Drivers tend to operate at a lower safety margin - Patten, Kircher, Ostlund, Nilsson, & Svenson (2006) found that YDs took longer to respond to a secondary detection task while driving in a simulator than more experienced drivers Saad, Delhomme & van Elslande (1990) found that YDs reduced their speed less when approaching a junction

conformers

always or nearly always comply with limit (60%);

violations

deliberate deviations from the those practices believed necessary to maintain the safe operation of a potentially hazardous system e.g. cross a junction knowing that the traffic lights have already turned against you Speeding and mobile phone use are universally the most common violations

defiers

exceed limits regardless (1%)

Mistakes

failure of planned acTIon to achieve intended consequence includes both failures of observaTIon and misjudgements UnderesOmate the speed of an oncoming vehicle when over taking Weak associaTIon with length of per-licence driving

Task Capability Interface Important concepts: Driver Capability

how competent is the driver

Task Capability Interface Important concepts: task demand

impact depends on driving situation: stalling the engine vs crashing at speed

components of task demand (TDe): Information input

interpreting the road environment ahead

Test-retest reliability

is much more difficult/rare measure to achieve since it requires retesAng the same sample Parker et al. (1995) aner 7 months retested 80/1600 original respondents Errors .69, violaTIons .81, lapses .75 indicaTIng relaTIvely high reliability across Ome High level for violaTIons indicates the disposiTIonal and therefore consistency of origin behind this behaviour

For an experienced driver with high Capability = ______ For an inexperienced driver with low Capability = _______

low TDi high TDi

components of task demand (TDe): response output

manoeuvring of car in appropriate fashion

self regulation older drivers

older drivers display more caution self-restriction or avoidance with increasing age

Lapses

potentially embarrassing, may be a source of inconvenience, but are not usually life-threatening More commonly reported by female and older drivers e.g. hit someone while reversing that you had not previously seen Paradox in that it requires parTIcipants to remember something that they may have forgot to do e.g. indicate before pulling out Unconscious errors are hard to remember since they are unconscious (Bjørnskau & Sagber, 2005)

compliers

reduce their speed at cameras and remain at or below the speed limit(13%);

manipulators

slow down for camera and then increase again(16%)

As TDe ↑,

spar capacity ↓ and driver becomes more vulnerable to error or the increased demands of emergency situations Using a mobile when driving on straight road remains within driver's capability but as soon as when TDe = responding to pedestrian crossing the road + demand of phone use TDe > DC resulting is loss of control

Components of TDe Contributed to by factors including

vehicle performance, instrument display characteristics, ergonomic match between car and driver, route, physical characteristics of the road and the presence and behaviour of other road users Can be reduced by technology such as lane assist Default method of reducing TDe is to reduce speed Important element is time to execute a behaviour/response The lower the required response window the greater the TDe Is contributed to by factors such as: physiological status, training, knowledge and experience Experience allows for the development of schemas to facilitate efficient navigation of scenarios previously encountered Enables preadaptation to anticipated changes in TDe in a top down fashion

Influencing perceived capability

younger drivers tend to have false perceptions on caability levels

Training

• Training relates to the assimilation of procedural knowledge - Driver training refers to an instructional programmes or set of procedures that relate to control of the car • A certain amount is usually mandated before a driving test can be attempted, and is usually performed by an ADI or non-professional supervisor • Evidence for the positive impact of training is much more compelling (Groeger & Clegg, 2007)

Developing a Driver Behaviour questionnaire Distinction between skill and style of driving Driver style/behaviour

- relates to how a driver chooses to drive and can be thought of as habitual responses to specific situation Eg, speed when opportunity arises

factor structure

3 factor model (Errors, aggressive violaTIons and ordinary violaTIons) was at least parTIally saTIsfactory in each of Finland, Greece, Iran, Holland, UK & Turkey (Özkan, Lajunen, Chliaoutakis, Parker & Summala, 2006) 2 factor structure (errors vs violaTIons) found across different groups including: professional drivers, motor cyclists, offenders, learner drivers, young women, and older drivers (de Winter & Dodou, 2010). Is predictive of collision involvement (Özkan, Lajunen, Chliaoutakis, Parker, & Summala, 2006) Error and violation scores 'predicted' collisions prospectively and retrospectively (de Winter & Dodou, 2010)

GADGET cells

4 Ls each divided into 3 columns - Knowledge and skills - describes what each driver needs to know at each level to drive competently and cope in normal traffic - Risk increasing factors - appreciation of the factors that increase risk e.g. hazard perception • Includes propensity for deliberate risk taking and underestimating of risks associated with driving • Training risk awareness is very different to car handling skills - Self-evaluation - awareness of skills and poor habits of aberrant tendencies

Relevance of mistakes in driving

65% of automobile accidents are wholly as a result and up to 95% of automobile accidents are at least partly as a result of human error" (Sabey & Taylor, 1980) Arguably the most frequently cited staOsOc is driver behaviour research - although the value of 80% is the variant of this usually proffered Implies a reducTIon in driver error would improve situaTIon Improvement of driver skills through training Reduction in workload through technological assistance

Understanding risk YDs

7 in-depth interviews with males (20-23) High awareness of risks, but notions of serious outcomes are not salient (Falk & Montgomery, 2007) 3 year longitudinal study assessing reckless driving, smoking and drinking Found adolescents are aware of the risks associated with their behaviour but modify their thinking about these risks in ways that facilitate continued participation in the behaviours (Gerrard et al., 1996)

Peer influence outside the car

Aberg et al., (1997); Gerrard et al., (1996); Gormley & Fuller, (2008) False consensus bias Those who exceed the speed limit overestimate the speed of other drivers Adolescents overestimate the prevalence of reckless driving among their peers and this overestimate predicts greater compliance with the perceived 'norm' This overestimation decreases with age Occurs in other behaviours such as smoking

Physiological vulnerability in YD

Alcohol has a greater impact on a young adult, all other things being equal, than on older adults (Box & Wengraf, 2013) YDs less likely to have built up a tolerance to alcohol Less experienced drivers require more conscious control which will be impacted more by negative effects of alcohol Due to an underdeveloped control system more likely to engage in risky behaviour and this is exaggerated by alcohol Younger drivers require more sleep, up to 9-10 hours per night, which is often not achieved due to poor sleep hygiene (Jackson et al., 2011) YDs more often drive under fatigued state - YDs more often drive late at night due to socialising Impact of sleep deprivation and alcohol combined becomes very detrimental

All levels of GADGET are important

All levels are interdependent and any mix of levels could be influencing driver behaviour at any one time • Higher level goals and motive should override skills and considerations at lower levels • Attempts to improve safety by concentrating on a single level have proved ineffective - Mandatory course on skid avoidance resulted in higher collision rates for participants, presumably due to over confidence (Christensen & Glad, 1996)

impression management

All self-report measures are potenTIally biassed by the desire to present oneself more favourably than reality Paulhus and Reid (1991) idenTIfied two disTInct factors Impression management - Self-decepTIon - is also an issue since an exaggerated sense of control and confidence in one's driving skills can lead to the objecOve safety margin being much smaller than the subjecOve safety margin (Summala, 1998) Driver impression management (lying) was negaTIvely related to the self-reported number of accidents and punishments, overtaking frequency, speeding, and driving aggression, and it was posiTIvely related to traffic rule compliance (Lajunen, Corry, Summala & Hartley, 1997) Af Wåhlberg (2010) found large correlaTIons between violaTIon scale and impression management (-.51) PredicTIve power was more than halved when social desirability was controlled for Lie scales must be included and used to correct associaTIons

Are official statistics any better?

Also problematic- definitions, lack of resources ,data protection issues (even before GDPR), variable period of data retention Records from police forces and insurance companies do not tend to record minor injuries Problematic since minor incidence important in their own right and can be precursors to more serious injuries Some groups are overrepresented Eg, older drivers are overrepresented because of their fragility- even minor collisions can result in injury

Design issues

Are not experimental designs with participants from the same population being randomly assigned to intervention vs control group • Self-selection to the intervention group is a huge issue and differences remain even if balanced for age, gender, socioeconomic status and educational attainment

Components of DC

As the balance of Conscious processes relative to unconscious processes increase so TDi increases (Fastenmeier & Gstalter, 2007) When learning to drive driving is primarily at the conscious level and is the reason why the task is so difficult Through skill acquisition/learning task execution switches from conscious to unconscious thus making the task less difficult

Wundersitz (2012)

Australian, 16-19 & 20-24 75% of teen driver crashes = driver error Most frequent errors related to decision making, followed by vehicle operation & perception Younger group made more vehicle operation errors while the older group made more perception based errors. Clearly indicates that perception based skills take longer to accumulate Males more likely to be involved in collisions involving speed or intoxication Over represented in single-car and loss-of-control collisions

Skills vs behaviour

Behaviour can be assessed using psychometric testing whereas skill is more difficult The Driver Skill Inventory (Lajunen & Summala, 1995) measures a subjective appreciation of ability through identification of weaknesses and strengths

Adolescent control system

Brain development involves Increased myelination, especially in prefrontal cortex and increased inter-neuronal connectivity Pruning and thinning of cortex The young brain does not develop fully until mid 20s Has led to speculation that the maturational lag in the prefrontal cortex leads to poor impulse control found particularly in young males

Important ed/training elements derived from Models

Calibration - Fuller (2008) - There needs to be self-reflection and evaluation in driver education as well as consideration of motives and attitudes (Gregersen, 1996a, 1996b) • Adequate psychomotor skill and physiological competence are not sufficient for safe driving; good decision making is also required - errors vs violations is an important distinction (Parker, Reason, Manstead & Stradling, 1995)

Task Difficulty Homeostasis

Central principle - drivers are continually making real time decisions to maintain perceived task difficulty within acceptable levels usually through speed adjustments Increases in TDe produced by darkness (Kilpeläinen & Summala, 2007) or mobile phone use (Liu & Lee, 2005) results in proportional decrease in speed Means that as TDe decreases with empty roads speed will increase (Broughton, 2005) TDe could also be decreased by increasing headway to car in front

Collision contributing factors

Children's Hospital of Philadelphia- N teen driver safety report, 2005-2011 75% of teen driver crashes = driver error 21% driving too fast for conditions - 21% not properly scanning 19% being distracted

Specific deficits older drivers

Cognitive impairment has a negative impact on situation awareness and processing speed and is linked to declined in cognitive, sensory and motor functioning Impacted ion angle collisions, crashes at intersections, turning, and changing lanes all incidents in which older drivers are over represented Older drivers focus on their own turning behaviour and structure of the roadway while middle aged drives focused on the changing behaviours of other road users Cognitive impairments is also linked to difficult adhering speed limit, failure to respond to stop signs, failure to monitor traffic, difficulties maintaining lane positioning and slower reaction time

young drivers stats

Concentrating on issues that are unique to YDs Watson, Watson, Siskind, Fleiter & Soole (2015) indicate that YDs who are serial speeders are also more likely to be involved in criminality - but the same is the case for all age cohorts so tells us nothing about the unique difficulties experienced by YDs Although 15-24 year olds account for 9-13% of OECD countries crashes involving YDs account for between 20-30% of all traffic fatalities YDs have a higher number of crashes at night and/or on weekends, when carrying similar aged passengers Alcohol and driving without a seat belt remain key factors in YD crashes Drug driving especially with cannabis is increasingly becoming an issue

Characterics of self-report measures in collision reports

Content of the response constitutes the data of interest Contrast with experimental techniques where reaction time or physiological response might tell us indirectly about the variable of interest Response is assumed to reflect respondents reality Individual must be aware of their own reality Have many advantages; less expensive and labour intensive than using instrumented vehicles or simulators, can provide more detail than observation, can reach large numbers of participants in quick time and data are relatively easy to analyse and interpret

Campaign effectiveness

Delhomme, De Dobbeleer, Forward and Simoes (2009) identify 5 factors which determine the effectiveness of a campaign - Types of tools to be used (e.g., posters, TV, radio, social media) - Campaign duration (e.g., weeks, months, years) - Social norms underlying target audience (e.g., perceptions about the acceptability of a behaviour) - External influences and environment in which the issue takes place (barriers to the behaviour, social control mechanism, road features) - Ensuring that intervention is based on theoretical model (Elliott, 1993

Drink Driving campaigns

Drink Driving remains a serious issue - From 2008-2012 of 867 fatal collision for which files were available, 38% cited alcohol as a contributory factor (RSA, 2016) • % of road deaths where driver/rider was over the legal limit was 20% • Comparable figure from the US in 2009 was 32% (Robinson & Holmes, 2011) • Comparable figure from the UK in 2014 was 13% (DfT, 2015) Wundersitz, Hutchinson & Woolley (2010) found campaigns effective if they were reinforced by other initiatives such as grassroots activities, police enforcement and other media messages • Elder et al. (2004) found the median decrease in alcoholrelated crashes was 13% • Vaa, Assum, Ulleberg & Veisten (2004) reported a Swedish meta-analysis that found a 14.4% reduction in collisions involving alcohol

How important is skill?

Driver training cannot be considered an effective crash counter measure Balance between safety and quality of life/better good for everyone Found that while a certain minimum level of skill is necessary for driving, higher levels of skill are no guarantee that the skill will be used appropriately Although the technical skills of manoeuvring a vehicle and positioning ir are important there is much more: Learning explicit and implicit rules, hazard perception- both real time and futuristic, response inhibition and healing with attention overload, higher order deciduous making

anxiety and confidence in ODs

Driving anxiety is common among older drivers with estimates of between 17-20% un the New Zealand population and frequently cited as the reason for driving cessation among canadians Confidence has been shown to decrease with age and functional decline and ultimately contributed to older drivers to stop driving Self-regulation based on lack of confidences as opposed to self awareness can be problematic and can lead to unnecessary driving restrictions instead of continuing to drive with the use of effective coping strategies Women report lower levels of confidence and more willingness to reduce driving or stop altogether As we get older, we are less likely to drive as we get olders (decline for both male and female but male plateau in beginning and lowers gradually weight age while female declines more severely) However, married women are more likely to drive compared to married males

impact of cessation in older drivers

Driving oneself intrinsically linked to social engagement and interactions Inevitably ODs will at some point be saved with the decision to stop driving Very hard decision for older drivers because causes a lot of difficulties in mental health Stopping driving is linked to negative outcomes such as reduced social interaction, poorer health, poorer quality of life, loneliness and depression

Education

Education relates to the communication of an established body of declarative knowledge and skills - Driver education refers to contemplative and value-based instruction of knowledge and attitudes relating to safe driving behaviour • Can come from ADI, course, parents or formal programme delivered through schools • Evidence for the positive impact of education is scarce (Ker et al., 2005) and could even make things worse (Williams & Ferguson, 2004)

errors vs violations

Errors (Lapses and Mistakes)- emanate from failure to act in manner the driver would have retrospecTIvely chosen - unintenTIonal violation - clear intenTIon to break rules of the road or social convenTIon High violaOon score related to high collision involvement in the past and in the future High violators not only more likely to run into others (culpable) but also to be hit by others (passive collisions)

Error 'development' with experience

Evidence from the YDs would support that fact that collision types first show handling and manoeuvring skills deficits followed by hazard percepTIon and then by decision difficulTIes

Content validity

Extent to which it covers the whole construct Requires invesTIgaTIon of the extent to which all the theoreTIcal constructs are measured by the items Has not been completed for the the DBQ or other main DB measures Unlikely to much of an issue since there are no obvious omissions that have been idenTIfied within the literature As behaviour changes in conjuncTIon with the switch to parTIally autonomous vehicles and level of distracOons induced by technology increases then this may change

unconscious processing

Few errors, high speed, little mutual interference Parallel in nature Limited range of contents, routines are relatively isolated and autonomous Appears to require little or no effort

Basic principles of driver ed/training

From Keating (2007) - Practice (driving) makes perfect - Solo driving gets safer with increasing age - Driving is a complex skill best taught by professionals - But note McKenna (2010) claims these are not governed by theory from the driver behaviour literature • Dr ed/training should go beyond the technical skills required to manoeuvre a vehicle (Twisk & Stacey, 2007)

Peer influence Inside the car

Gardner & Steinberg (2005) Simulated driving recorded in the presence or absence of peers In the presence of peers, participants: Took more risks - Focused more on benefits than costs Made riskier decisions Peer influence stronger in adolescents (13-16) and youths (18-24) than in adults Negative impact of age related peers (Simons-Morton et al., 2011) Distraction created by passengers exacerbated by poor inhibitory abilities Have little awareness of the difficulties they create for driver Have appreciation of risk but just think it doesn't apply to them

Changes in population of <60s

Globally in 2018 proportion of population over 60 is 11-12 % but this will almost hubble ny 2050 to 21022 Proportion of older drivers (<60) is highest in developed countries In 2015 japan 26.6% and US 14.9% In ireland, driving oneself was identified by 76.1% of 50+ as the most frequently used form of transport Private cars remain the preferred transport option for o;lder people who travel more than previous generations UK population is predicted to rise substantially by 2040 Percentage of people who are going to experience difficulties driving is going to increase with this increase of population Fatality rate is highest in 85+ year drivers Partially because of fragility However, there is a decline in crash involvement the older people get

Psychosocial functions of driving

Gormley & Fuller, 2008 When asked about the consequences of a dangerous driving scenario older drivers likely to list death first as the consequence Younger drivers likely to list it as the third or fourth consequence Implication is that accurate risk information is available but not immediately accessible Attract attention Identify with adult status Attract status through fast driving, demonstration of mastery and taking risks Driving as pleasure activity: A to A

Interventions: young driving

Graduated Driving licensing Specified learning stage Staged introduction of driving privileges for newly qualified drivers e.g. restrictions on night time driving and driving with peers as passengers and lower alcohol limits Russell, Vandermeer & Hartling (2011) found that reductions in crash rates were seen in all jurisdictions and for all crash types But most effective combination of measures remain elusive Black box telematics can record speeding, lane departure and harsh braking and risky driving Used by insurance companies to reduce premiums if customers agree to drive within stated limits All users are self-selecting therefore making evaluation difficult Parents (Cattan et al., 2008) Parents' driving styles have an impact on teenagers' car safety behaviour and, in particular, seatbelt use Parental monitoring of the learning-to-drive and post-driving processes has also been found to positively influence their child's driving

GADGET (GDE)

Guarding Automobile Drivers through Guidance, Education and Technology (Hatakka, Keskinen, Gregersen, & Glad, 2002) - An attempt to go beyond the skills view of driving - a psychomotor task which is trainable and includes knowing and following traffic rules - Offered a theory-based goal structure for improving driver education - Describes important goals and contents - Examine learning methods to improve factors that directly impact on the safety of YDs

Evidence for TDiH (task difficulty homeostasis)

Horst (2007) found that when drivers were asked to brake hard to stop in front of simulated vehicle, time to stop was consistent independent of approach speed Stradling et al. (2008) found wide variation in normal speed for a single carriageway rural road and the speed that would put them right at edge of safety margin Consistent relationship between the two speeds with the later representing on average a 14% increase Feeling of risk (proxy of TDi) was independent of speed indicating a consistency of TDi across drivers

Distal

Impact of stimuli can start before action

Importance of distance travelled in collision involvement

Important DV (dependent variable) since it is measure of risk exposure and influences likelihood of being involved in a collisions Increased mileage ⇒ increased collision involvement Important to take into account when looking at statistics like demographics involved in accidents ( ex, do males drive more than females which is why they get in more accidents?) Relationship is not linear but rather a negatively accelerating curve with smaller increases in collision rate at higher levels of mileage High mileage drivers tend to drive on safer roads such as motorways where collision rate/unit distance travelled is smaller this reducing accumulated exposure risk All kilometres are not the same Impact of mileage sends on: type of road, time of year, time of day, purpose of journey Ex, highways are safer than country roads Winter is problematic bc cold, peak of summer is problematic bc lots of people are travelling Time of day- when people go to work and leave from work Weekend later hours peak (driving intoxicated or sleep deprived) Ignoring exposure will increase error variance and reduce ability to detect relationship between physiological variables and collision involvement Has been used to try explain what men are more collision involved than women- but not even when this is controlled for men are still overrepresented in collision stats

Evaluation of Dr Ed/training

Is rarely about the educational process with some exceptions e.g. - Katila et al. (1996) students produced different views on the important factors for safe driving in slippery road environments - HERMES project (2010) - used a coaching method to support the learning of YDs to make them more responsible for their learning and subsequently their driving • Was suited best to more mature female students with higher education

Dementia

It is estimated that 1/9 Americans suffer from Alzheimer's disease, a proportion which increases with age When engaged in complex tasks they fail to adapt their decision-making strategies leading to a higher frequency of random decision making Are cognitively slower With MCI, the overall impact on collisions is not clear Between 2.5-5 more likely to be involved in a collision as drivers compared to those without dementia Fraade-Blanae et al (2018) indicates reduced collusion involvement compared to age matched controls but this likely due to decreases in mileage and other self-regulatory behaviour including avoidance of complex driving scenarios Research on dementia driving is very ambiguous because it is hard to test in an experimental setting

GADGET level 2: Mastery of traffic situations

Knowledge of how to drive in certain traffic situations • Be able to anticipate and adjust your driving in response to the dynamic traffic environment • Knowledge of traffic rules, hazard perception

Using principal components analysis (EFA) behaviours were grouped into 3 basic types:

Lapses Errors Violations (Reason et al, 1990; Meadows, 1994; Parker et al, 1995a,b) Labels used were an attempt to differentiate the latent constructs - usually difficult and can cause confusion Errors later became known as Mistakes with term Error being used to refer to both Mistakes and Lapses Important to remember the limitations of EFA Content validity Garbage in garbage out

'Maladapted' control system

Likely consequence is the participation in risk related behaviours BUT only a subset of adolescents show such risky behaviour (O'Callaghan, Kelly, & Gormley, 2015; Bjork & Pardini, 2015) Leads to individual differences similar to e.g. IQ

Fuller et al. (2008) found 4 types of driver

Low risk threshold (40%) generally comply with speed limits High risk threshold (14%) speed frequently in addition to other dangerous driving and have more convictions Opportunistic (23%) will speed if they think it is safe to do so Reactive (23%) speed when angry, annoyed or under time pressure

Health Belief Model 2

Main motivator for people to preserve or protect their health is to avoid negative health related outcomes and is influenced by 6 concepts (Carpenter, 2010) - Susceptibility and Perceived seriousness combine to give perceived threat - Perceived barriers and benefits combine to form a cost-benefit analysis • Carpenters meta-analysis found perceived benefits and barriers to action were the greatest predictor of behavioural outcomes In a anti-drink driving campaign the message may emphasize the perceived barriers i.e., the likelihood of being caught by the police while being over the limit and the consequence i.e., loosing your license

Development of the DBQ

Manchester Driver Behaviour Research group conducted numerous surveys addressing different driving scenarios. How often have you: Misread the signs and exit from the roundabout on the wrong road On turning let nearly hit a cyclist who has come up on your inside Cross a juncTIon knowing that the traffic lights have already turned against you

Assessment in ODs

Mandatory assessment tends to related to eyesight resting In UK self-declaration of medical and visual fitness to drive Value of mandatory assessment is questionable No single cognitive screening instrument has sufficient validity and reliability to identify even the majority of unsafe drivers On road assessment remains the gold standard with high validity but can suffer from poor reliability due to the bvariabil;ity in the driving scenario

conscious processing

Many errors, slow, mutual interference between simultaneous processes Serial in nature Great range of contents, great ability to relate different conscious contents to each other Requires effort

Effectiveness of fear campaigns 2

Messages that induced fear but without feasible recommendations produced resistance and rejection of the message (Witte & Alle, 2000) • Aldoory & Bonzo (2005) suggested 3 outcomes from fear appeals - Danger control response - take action to avoid threat - Individuals become too fearful to take action - Ignoring the message because the threat is irrelevant/insignificant • Nielsen & Shapior (2009) found evidence of suppression and avoidance of threatening information

self-regulation of driving in ODs

Modification of driving habits in response to declines of functional abilities of the core of successful self-regulation Often self-regulation is spontaneous but antecedents? Age Related functional declines have been associated with reduced mileage Age itself prophecies reduced mileage Influence by personal interests, improving or maintaining self-esteem and self-definition

Wordload olderdrivers

ODs require more effort to perform cognitively demanding tasks but effort is determined by perceived task difficulty rather than objective task difficulty Increased effort measured by changes in eye movements and visual scanning patterns ODs engage in increased use of stereotyping in complex situation to reduce cognitive load which can lead to perceptual errors Increased workload relate to : slower cognitive processing, delayed response times, greater levels of distraction and therefore increases crash risk ODs are susceptible to in vehicle distraction such as mobile phones, passengers, and eating and drinking

Issue of collision involvement

Period of remembering a clear issue 14% of people involved in road collision did not remember the incident 1 year later Maycock found over 3 year period the level of forgetting was around 30% p/year Even severe collisions be be routinely forgotten

Overall effectiveness

Phillips, Ulleberg & Vaa (2009) looked at 437 effects extracted from 228 international studies across 14 countries from the previous 30 years - Reduced the number of road incidents by approximately 9%; > increased seatbelt use by 25% - Reduced speeding by 16% - Increased yielding behaviour by 37% - Increased risk comprehension by about 16% • Phillips, Ulleberg & Vaa (2011) looked at 119 effects extracted from 67 looked at the features that were effective in reducing crash frequency - Shorter duration (less than one month) - Personal communication - Roadside delivery, use of roadside media, or delivered in proximity to the behaviour occurring • Phillips, Ulleberg & Vaa (2011) looked at 119 effects extracted from 67 looked at the features that were effective in reducing crash frequency - Combined emotional/rational message has a stronger influence than a purely rational message - Accompanied by enforcement - Combined with mass media

Driving is a complex process It Requires

Precise coordinated limb movements to manoeuvre the car Continual appreciation of Visual stimuli to judge location, speed, trajectory and intention of other road users Auditory stimuli within and outside your vehicle and tactile stimuli from vehicle Safety of road environment and own vehicle Knowledge of: Explicit and implicit rules of road Inherent difficulty of driving and one's own ability to drive

Cognitive Impairment in OD

Prevalence of mild cognitive impairment is estimated to range between 5-36% for european countries Impact of cognitive impairment is unclear Vernon et al (2002) looking at Utah drivers found those with self-declared cognitive impairment were 5 times as likely to be involved in an at fault collision compared to age matched controls Fraade-blanae at al (2018) found an adjusted hazard ratio of collision involvement among US drivers diagnosed with dementia of .56- likely due to self regulation Cognitive impairment lead to greater rate of increased driving difficulty and reduced driving in challenging situations Self-awareness of functional decline seems to encourage adaptive behaviour in drivers with cognitive impairment, particularly in situations that they find more demanding

Issue of Age of novice US drivers

Pushing solo driving toward older age should be one of the key goals (Keating, 2007) - But in 8 US states you can get a learner's permit at 14/14.5, in 39 states a restricted license at 16 or younger, and in 34 states a full license by 17* or younger (Verywellfamily, 2021) • In the case of South Dakota you can get a learner's permit at 14, a restricted license at 14.5 with full license at 16!

Are older drivers more dangerous

Rarely involved in risk taking but rather fail to respond appropriately to traffic situations intending to be safe Ex, rear ending person in front or driving through a red light Fue to fragility, any crash has a greater impact on them Crash rate increases substantially after 75 years old physiological decline starts mostly at 75 a lot more Causality crashes are significantly more prevalent amongst 80+ year olds

approaches in ODs

Reactive approaches to encouraging self regulation or driving cessation are associated with more negative consequences than planned processes Planning fives older drivers a feeling of control over the process and means that they are likely to have a better quality of life after giving up driving

Impact of age and driving experience on collision and liability from licensure

Reduction of crashes/ 1000 drivers was about 50% in first 8 months in Norway (Sagberg, 2000) In Germany, reduction of 50% in first 9 months (Schade, 2001) In Canada, 41% during first 7 months (Mayhew et al., 2000)

Improvements in vehicle technology

Reductions in ODs fatality states have been attributed to ompaced vehicle design 'collision forgiveness' Drivers are now more protected Driver assistance technology has also improved the experience of older drivers through avoidance of rear ending and unintended lange departure and blind spot detection Although ODs have reported difficulties using new technology in cars, they have been willing to engage with them

YD issue in Ireland

Relative fatality rate of 2.6 (somewhat of an outlier), compared to 2.3 for France and EU average of 1.7 (ERSO - Traffic basic facts, 2017) Irish data from 2013, other countries was 2015 2016 data have the value as about 1.7, 3rd worst Second highest rate of car related deaths of approx. 85% compared EU average of approx. 65 and UK approx. 58% Largest % killed on a Sun - 28; nearest is Cyprus at 18 while EU average is 10

Self awareness in ODs

Self-awareness is a key factor in driving self-regulation Those lacking self awareness pose risk to self and others Self awareness may be negatively impacted by disease processes with an inevitable knock on effect on self0regulaiton Self rating of ability failed to product older drivers functional performance related to cognitive, visual, and physical abilities Drivers with established cognitive impairment avoid more situation than those with questionability or no cognitive impairment because professionals have told them how to combat and recognize these cognitive impairments Provided lower self ratings indicating that they understood that they were experiencing negative performance effects Improved though various tools Eg, MTRI Driving Decisions Workbook drivers using it reported increased awareness of cognitive and operational deficits, which triggered self-regulation processes, with 14% becoming aware of changes in their abilities and 25% reporting intention to change their driving behaviour in order to mitigate the risks associated with age-related declines

Best option: collision involvement measurement

Self-report up to a max of 5 years complimented where possible with state records

vision OD

Status visual acuity (ability to see clear images), contrast sensitivity (light and dark) and depth of visual field all decline with age and is likely to impact on driving Hard time with night time driving Declining contrast sensitivity linked to increased crash rism With every 13 years that passes from age 20, drivers require twice as much light to drive safely 72 year olds require 16 times as much light as a 20 year old Age-related eye diseases: cataract\facts, glaucome, diabetic retinopathy, and macular degeneration pose significant challenges for older drivers Darkness is not the only difficulty as wet weather condition s produces difficulty through rain, sprat, and dim ambient lighting Drivers with lower visual function driver fewer miles Declines in contrast sensitivity have been linked to driving cessation- vision is something that we are very self-aware about other than other problem that occur when driving as we get older Impact on overall crash risk remain unclear because it is hard to experiment in a research setting

'Measuring' distance travelled

Suffers from usual self-report difficulties but usually the only option available in large population based studies Potentially between that self- report of collision involvement Collision involvement is a rare event whereas there is a certain degree of regularity relating to distance travelled on a daily, weekly, and monthly basis allowing for estimation Most likely to be distorted for low and high distance drivers Objective measures are better but not always feasible

Task difficulty

Task difficulty = task demand - driver capability TDi= TDe - DC Since capability is reasonably constant, as TDe increases so TDi increases

Accommodating ODs during design

Tendency for vehicle systems to be designed without sufficient consideration of ODs opinions, capabilities, and needs leading to increases in complexity of driving Designs in cars are also made for a middle age adult males (worse for short people and females) Only effective collaboration between developers, manufacturers, engineers, researchers, governments and road safety safencies, can automate vehicle technology be completement to full effect in such a way that it benefits the older driver population Designing for older people, and developing intuitive and user-friendly systems would benefit the entire driving population

Measuring collision involvement

Term Road Traffic collision (RTC) is much preferable to rad traffic accident (RTA) Accident makes it seem as though no one is to blame for the collision vs collision opens the possibility of someone being to blame Only situation where no one is to blame- a wild animal Collision involvement is an important DV Eg, what are the antecedents of being collision involved Reporting of collision involvement can be problematic McGwin, Owsley and Ball (1998) found poor equivalence between self-report and state records for collisions involvement among older drivers We are not particularly good at remembering previous accidents/collisions Boufous et al (2010) found good correspondence between police records and self-report of younger drivers

Evaluating Driver ed/training

The ultimate measure is the number of KSIs (killed or seriously injured), particularly in the first 6 months post licensing • Other measures are also important (Hatakka, 2003) - learner's satisfaction with training - immediate learning effects (passing rates on examination and quality of mistakes on examination) - novice drivers' attitudes and their behaviour in traffic when fully licensed (offenses and other measures of driving style and the amount of driving) • Other measures are also important (Hatakka, 2003) - Safety compromising results of driving behaviour (number and severity of collisions) - The time of collisions (during learner stage, intermediate stage or when fully licensed)

Effectiveness of fear campaigns

Wundersitz, Hutchinson & Woolley (2010) found that positive emotional appeals were more persuasive for males than fear appeals, whereas the opposite was true for females • Fear appeals have been shown to be less effective among those who do not feel vulnerable (Cismaru, Lavack, & Markewich, 2009) - They also showed that campaigns that outline coping mechanisms invoke greater change than fear alone

Social Norms Theory

suggests that people base their behavior partly on what they consider appropriate and socially acceptable Often perception is flawed meaning that false information is impacting on behaviour - Gormley & Fuller (2008) found that although reported risky behaviours such as speeding were quite prevalent they were less prevalent that participants believed them to be • Potentially generates a positive feedback loop which reinforces behaviour • Also the possibility that retrospectively aberrant drivers identify high descriptive norms to justify their own behaviour - Forward (2009) found that the use of descriptive norms significantly increased viability of TPB model for speeding/dangerous over taking

Construct validity

the extent to which variables measure what they are supposed to measure Smotri & Guarnieri (2016) - differentiated between younger and older drivers Importantly distinction between errors and violations is universal (Lajunen, Parker & Summala, 2004) and sample invariant (de Winter & Dodou, 2010)

Irish system - GDL light

• 12 hours of Essential Driver Training delivered by an Road Safety Authority EDT approved Instructor - Each session must be stamped by your Approved Driving Instructor - EDT must span a minimum of 6 months • Lower alcohol limits (.02 rather than .05) and double penalty points e.g. for speeding, for learner and restricted drivers • N plates must be displayed for 2years post successful test

Theories of social persuasion

• Addresses the perceived social norms of the corresponding population e.g., YDs • Social Norms Theory (Yanovitzky, 2004) - people tend to act in accordance with perceived group expectations and behavioural norms - Need to feel part of reference group - Social comparison processes - In-group pressure to conform - Availability of reference groups norms

Accompanied driving

• Aim is to offer YDs a long protected practicing period • Different jurisdictions vary in terms of characteristics of accompanying person - In France the accompanying person had to receive instruction from a driving school - In Sweden and Norway limited restrictions on who could do the accompanying • Only clear success story was from Sweden (Gregersen et al., 2000)

Place of education in road safety

• Arguably the mostly used term in road safety is the three E's: Education, Enforcement and Engineering - Julien H. Harvey, who was then director of the Kansas City Safety Council, gave a presentation in 1923 during which he presented a drawing of a triangle with sides labelled with one of each of the Es (Groeger, 2011)

Road Safety Campaigns

• Campaigns are an important component of RS • Used to: increase awareness of traffic laws and new legislation (e.g., metrication conversion of miles to kilometres - Jan 2005, Ireland), safe driving practices (heed your speed) and education about issues relating to RS (driver fatigue) • Versatile since they are scalable, temporally variable, can be tailored to specific audience and can adopt to media trends e.g. social media

North American vs European approach

• Canada and US has favoured a Graduated Licensing programme - essentially based on long-lasting training (Shope, 2007) - Drivers progress through phases and can only reach next stage once competency has been demonstrated - Have to assume that its prevalence is to do with age • Europe has favoured Driver Education and testing regime developed according to theoretical models

Fear based campaigns

• Confront audience with negative consequences of risky behaviour in a graphic way to induce negative affect in the form of shock/fear/distress • May rely on graphic imagery of crash footage or injuries - (http://youtu.be/Z2mf8DtWWd8) NOT FOR THE FAINT HEARTED - after about 2:22/5:21 is gets VERY graphic! Is from Transport Accident Commission Australia who arguably lead the way in such advertising

GADGET L3: Goals and Context of Driving

• Decision level - navigational and planning tasks • Trip-related goals - why you are driving to where, when and with whom - With whom particularly problematic for YDs - when a YD drives with peers their crash risk increases while when an OD drives with peers their crash risk goes down (Williams, 2003) • Will include a decision about whether to drive based on: level of impairment, stress and tiredness levels

Driver Licensing models

• Driver licensing systems differ on a number of dimensions: licensing age, whether education in mandatory, the curriculum of such education and how this education is delivered e.g. professional or nonprofessional (OECD, 2006) - Professional instructors deliver both theory and practical training - Education is based on a curriculum generated by the relevant authority Driving school education has set goals and content for training, organised and structured teaching, emphasis on feedback and integration between practical and theoretical instruction

Theory of Planned Behaviour

• Foolspeed five-year road safety campaign conducted by Road Safety Scotland to reduce excessive speeding - Produced a TV advert to address each component of theory • Stead & Eadie (2007) - Some ads based on the TPB components were less effective - Campaign demonstrated that it was possible to design convincing and effective campaign materials based on the TPB

Application of GDE framework 2

• If the motivational level (L4) does not impart a safety ethos no level of appreciation of how to navigate traffic will reduce collision risk for the driver • Driving training is good at ensuring competence at Ls1-2 but not Ls3-4 and only education can address the higher levels, especially L4 - few courses do this properly, if at all

GADGET L4: Goals for Life and Skills for living

• Importance of Personality factors such as impulse control, attitudes, gender, age, group affiliation and peer influence, and importance of cars and driving as part of one's identity • Relevance of physical and cognitive capabilities - driver needs to be aware of potential limitations to ensure negation of potential negative effects

Multiphase Driver Education programmes Finland

• In Finland after licensure drivers have to follow a second phase of driver training within 6-24 months - Takes place in an authorized driving school and has 3 parts • Analysis of personal driving skills and style • Track training units • Classroom training - But results were not encouraging with no significant decrease in collisions (e.g. Laapotti, Keskinen & Rajalin, 2003) - Arguably because participants were not self-selected for intervention (compare Willmes-Lenz, Prücher, F & Grossmann, 2010)

Transtheoretical Model of Change 2

• In order for permanent behaviour change to occur individual must go through each of the stages successfully and completely • Stages are fluid such that individual is capable of moving backwards and forwards through the stages • Pre-campaign research required to determine the dominant stage within the target population Drivers are aware of the dangers of speeding but nevertheless continue to speed therefore campaign should address the contemplation stage by encouraging self reflection on the antecedents and potential consequences of their speeding • Additional campaign elements should guide individuals along the path to full acceptance and maintenance of driving within the speed limits

GDL (Graduated Drivers License)

• Is designed to overcome inexperience and learn through doing but does not address deliberate risk-taking (Waller, 2003) - She further recommended - Initial experience should occur under low-risk conditions - The system should be based on extended supervised practice and the delay of onset of solo driving - Learning conditions should gradually move from simple to more complex - The system should punish beginning drivers' deliberate risk taking using harsher penalties

GADGET Level 1 - vehicle manoeuvring

• Is the lowest level and is subservient to the 3 higher levels • Emphasis is on perceptual and motor skills required to control the vehicle • Includes basic skills such as: smooth take off, gear changing, maintaining road position and braking • More complex skills: over taking; evasive manoeuvring; awareness of space around car, including blind spots and the successful use of mirrors

GDL evaluation

• Karaca-Mandic and Ridgeway (2010) argues that GDL policies reduce the number of collisions of 15- to 17-yearolds by limiting the amount of teenage driving rather than by improving teenage driving - Teen driving quality does not improve post GDL exposure and does not make teens better drivers in later years • GDL programmes are largely effective regardless of specific details (Hedlund, Shults & Compton, 2003) but primarily due to extended learning period and delay of solo driving 2

Health Belief Model

• Main motivator for people to preserve or protect their health is to avoid negative health related outcomes and is influenced by 6 concepts (Carpenter, 2010) - Susceptibility to the consequences of an action (e.g., perception of likelihood of crashing when speeding) - Perceived seriousness of consequences of action - Perceived barriers (costs) that decrease the likelihood of action - Perceived benefits that increase the likelihood of action - Confidence in ability to take action - Internal and external cues/motivators to affect likelihood of action

GDL evaluation 2

• McCartt, Teoh, Fields, Braitman, and Hellinga (2010) found influential elements were: night-time and passenger restrictions, age delay for acquisition of both the learning permit and the full licence - Also found the programmes with more restrictions were more effective • Not all results are positive - Vanlaar et al. (2009) found no impact on the relative fatality risk among 17, 18 and 19- year-olds

Application of GDE framework

• Risk awareness is as important as technical skills • Curriculum should cover all 4 levels • Driving is related to the driver's goals, motivation as well as skills in self-regulation • Active learning, making use of the learner's own experiences are needed to reach the higher levels • Training of self-evaluation should be included

Behaviour change theories

• Theory of planned behaviour - intentions to carry out behaviours predict behaviours and are based on (Ajzen, 1991) - Attitudes/beliefs about a behaviour e.g., some people find speeding to be thrilling - Subjective norms - how significant others perceive the behaviour (also known as injunctive norms) - Perceived behavioural control - Additional predictors • e.g., Descriptive norms and influence of habit (Donald, Cooper & Conchie,