Traffic Psychology and Sustained Mobility
Lecture 1 Theories and Models
Theories and models;
Engineers make mathematical descriptive models. Psychologists make conceptual models. A good
psychological theory must be descriptive and should predict. If it can predict then you can test it.
Driving is generally self-paced, can position how fast we go. It requires sustained attention, hazard
identification and motor-coordination. It places drivers seconds from death but is it forgiving. Models
apply to mobility. Eye-hand coordination is required. Environment is forgiving to voorkom accidents
and also other drivers want to voorkom it.
Driving kills one million people every year! A lot of people die! Young people mostly.
Loads of theories are proposed and none are widely accepted in the real world. Theories are used in
traffic.
Skill models -> a driver perceptual and motor skills are what makes them safe (reaction time, vision
and training). Low level model, concerned with reaction time, what we perceive. What are you able
to do and what is required. Tasked demand what you have to do and driver skills is what you have to
do. Crashes occur when Task Demands exceed Driver Skills.
Older drivers die more because they are vulnerable and fragile. Young drivers get more injuries, they
take more risks, not as experienced. They misjudge situations.
The model is too simplistic. Only a weak relationship between the attributes to accidents.
Psychomotor and perceptual variables are less important than biographical and exposure factors.
There are other properties more important, like exposure and experience. Reaction time is not very
important but mileage, martial status, socioeconomic status are most significant.
Attitude theories -> Theory of Planned Behaviour -> concepts are attitude
(what I say I think), subjective norm (what I think other think), perceived
control (how much control I think I have over my behaviour), intention
(what I say I will do) , behaviour (what I actually do).
Attitude to behaviour, lead to norm and now intention and this leads to
behaviour.
Big problem = very good in prediction intention but in the end we want to know about behaviour.
There is a big intention-behaviour/attitude-behaviour gap. When it comes to behaviour it does not
predict very much. We humans are not as rational as he implies, we do not plan all our behaviour.
Attitudes and beliefs about speeding did not change but the behaviour is adapted! Behaviour is good
when changed because that causes accidents in the end.
It is a popular theory because it is an easier way to do this research. Questionnaires are handy.
- Intention is not behaviour!!!
Attribution bias -> it is common sense.
Strong habits are more important in forming behaviour. Intentions play a role in building up habits.
*Intentions were related to behaviour when habit was weak.
Utility theories -> maximize gain and minimise lose! Rationally come to a conclusion. Unbounded
rationality?
Popular in modelling and in economics. Assumes rationality, think over all factors that matter. It tries
to explain homo sapiens, process in very slow. For a lot of behaviour in traffic it is not really working.
Rational approach does often not work.
,Risk/Motivational theories -> takes into account motives. Lie detectors! Positive or negative stimulus
can have a similar response so it makes it difficult. Safety, pleasure, time etc.
Risk Homeostasis Theory is important! The main aspect -> we strive towards feeling a subtle level of
risk all the time, the target risk. Society but mainly the individual influences the target risk. If it is too
risky than want more safety and not enough risk then want it more unsafe, behavioural adaptation to
it, go towards to target risk. Winter tyres, feel more safe, take more risks.
Problem with RHT -> difficult to test, not testable predictions. Continuous monitor risk, constant
perception of risk, so it is difficult to falsify. Also safety measures are undone by unsafe behaviour but
that is not the case!! It also implies that the population learns from crashes while only a small
proportion experience a crash. Difficult to learn from accidents that happen to others. Accidents are
rare so it is difficult to learn from it. It is also not specific through what pathways and over what
timeframe this occurs. Not clear where the risk monitor is located and what it uses as parameters.
Message -> we adapt to changes in the environment. Behavioural adaptation, the opposite direction
of intended measures.
Risk Allostasis Theory, RAT (Task Difficulty) -> risk plays a role but it is not the main thing that is
monitored, somatic marker, emotions, cognition causes the bodily response. Feeling of risk is the
central motivator and an indicator of perceived task difficulty. There is a preferred range. Evaluate
situation and have a response to it. Fuller moves away from risk and task demands and difficulty,
respond to it. Threshold, at a certain moment there will be a response. People who lack emotions
should be irrational then but can perform well in tests!
Allostasis vs Homeostasis -> dynamic (changing motivational influences) vs fixed target level of task
difficulty or risk. Motivation has a influence in allostasis.
Capability and Task Demands = central element in RAT!!!
RAT has testable hypothesis -> evaluate task difficulty, risk, loss control by speed. Threshold is passed
eventually, more risk after that. Not linear relationship.
Zero-risk theory -> hardly experience risk, when risk then it is a warning. Threshold is the main thing,
there are margins. It is a learning theory. Criticized because risk is not continuously monitored so
how know that threshold is past. How the extra motives function is vague. Not easily tested! Need to
learn and create safety margin, but requires too many contingencies.
Risk/threat avoidance theories -> Risk is aversive and avoided. Behavioural contingencies are learnt.
What is risky depends on experience.
*RISK -> the probability of an outcome * the consequence of that outcome. Risk is not the same for
everybody. We are at risk but do not experience it every time!! The calculation requires the
knowledge level; effortful and biased.
Risk monitoring -> target level of risk; target feeling of risk/task difficulty or zero risk.
Safety Margin Model -> threshold model, motivated to be comfortable (threshold). It is time based:
keep constant, time to collision and line crossing (Summala). It is very unconscious, do automatically
and not think about it. Safety margins are monitored; personal space.
2 categories of perception:
-continuous/constant perception = RHT, Task Difficulty Homeostasis, RAT
-threshold perception (threshold past and then aware) = Zero Risk Theory, Threat avoidance theory
and multiple comfort zone model (Summala)
Michon’s Model (Janssen’s) => performance and decision making.
Rasmussen – human performance. We perform on different levels, skills (automated, no effort),
rules (looking for similarities and copy appropriate behaviour, if and then, pattern matching and
recognition, more effort and time) and knowledge (bounded rationality and thinking, don’t do it very
often, it is slow and takes effort and time).
, Behaviour starts at a high level but we want to operate as a skill level. By automating it, it becomes a
lower level and then it becomes easier and manageable. This in an energy efficient way of dealing
with the world.
Hierarchical decision making -> strategic level (long time, general plans), manoeuvring level
(seconds, controlled) and control level (automatic behaviour).
Research: measures on these different levels. Low/Control level -> Longitudinal control(speed),
lateral control (in lanes) and reaction time. SDLP = standard deviation of lateral position -> when to
much than problems, want a low SDLP, when it rises then more risk.
Manoeuvre/intermediate -> car following and gap acceptance ; when more impulsive behaviour
(XTC), then accept smaller gaps, they are willing to take more risk.
Strategic/top level -> route choice or mode of transport.
Conclusion: Skills are necessary but only to a point. Change behaviour, not attitudes (easier and more
effective). Do not expect rationally – people are not calculators and emotions play a role. A lot of
behaviour is habitual/well learnt (change is early). There are critical decision points (threshold
crossed). The best way to change behaviour is to change the environment in which people operate,
not to try and change the person directly.
Lecture 1 Theories and Models
Theories and models;
Engineers make mathematical descriptive models. Psychologists make conceptual models. A good
psychological theory must be descriptive and should predict. If it can predict then you can test it.
Driving is generally self-paced, can position how fast we go. It requires sustained attention, hazard
identification and motor-coordination. It places drivers seconds from death but is it forgiving. Models
apply to mobility. Eye-hand coordination is required. Environment is forgiving to voorkom accidents
and also other drivers want to voorkom it.
Driving kills one million people every year! A lot of people die! Young people mostly.
Loads of theories are proposed and none are widely accepted in the real world. Theories are used in
traffic.
Skill models -> a driver perceptual and motor skills are what makes them safe (reaction time, vision
and training). Low level model, concerned with reaction time, what we perceive. What are you able
to do and what is required. Tasked demand what you have to do and driver skills is what you have to
do. Crashes occur when Task Demands exceed Driver Skills.
Older drivers die more because they are vulnerable and fragile. Young drivers get more injuries, they
take more risks, not as experienced. They misjudge situations.
The model is too simplistic. Only a weak relationship between the attributes to accidents.
Psychomotor and perceptual variables are less important than biographical and exposure factors.
There are other properties more important, like exposure and experience. Reaction time is not very
important but mileage, martial status, socioeconomic status are most significant.
Attitude theories -> Theory of Planned Behaviour -> concepts are attitude
(what I say I think), subjective norm (what I think other think), perceived
control (how much control I think I have over my behaviour), intention
(what I say I will do) , behaviour (what I actually do).
Attitude to behaviour, lead to norm and now intention and this leads to
behaviour.
Big problem = very good in prediction intention but in the end we want to know about behaviour.
There is a big intention-behaviour/attitude-behaviour gap. When it comes to behaviour it does not
predict very much. We humans are not as rational as he implies, we do not plan all our behaviour.
Attitudes and beliefs about speeding did not change but the behaviour is adapted! Behaviour is good
when changed because that causes accidents in the end.
It is a popular theory because it is an easier way to do this research. Questionnaires are handy.
- Intention is not behaviour!!!
Attribution bias -> it is common sense.
Strong habits are more important in forming behaviour. Intentions play a role in building up habits.
*Intentions were related to behaviour when habit was weak.
Utility theories -> maximize gain and minimise lose! Rationally come to a conclusion. Unbounded
rationality?
Popular in modelling and in economics. Assumes rationality, think over all factors that matter. It tries
to explain homo sapiens, process in very slow. For a lot of behaviour in traffic it is not really working.
Rational approach does often not work.
,Risk/Motivational theories -> takes into account motives. Lie detectors! Positive or negative stimulus
can have a similar response so it makes it difficult. Safety, pleasure, time etc.
Risk Homeostasis Theory is important! The main aspect -> we strive towards feeling a subtle level of
risk all the time, the target risk. Society but mainly the individual influences the target risk. If it is too
risky than want more safety and not enough risk then want it more unsafe, behavioural adaptation to
it, go towards to target risk. Winter tyres, feel more safe, take more risks.
Problem with RHT -> difficult to test, not testable predictions. Continuous monitor risk, constant
perception of risk, so it is difficult to falsify. Also safety measures are undone by unsafe behaviour but
that is not the case!! It also implies that the population learns from crashes while only a small
proportion experience a crash. Difficult to learn from accidents that happen to others. Accidents are
rare so it is difficult to learn from it. It is also not specific through what pathways and over what
timeframe this occurs. Not clear where the risk monitor is located and what it uses as parameters.
Message -> we adapt to changes in the environment. Behavioural adaptation, the opposite direction
of intended measures.
Risk Allostasis Theory, RAT (Task Difficulty) -> risk plays a role but it is not the main thing that is
monitored, somatic marker, emotions, cognition causes the bodily response. Feeling of risk is the
central motivator and an indicator of perceived task difficulty. There is a preferred range. Evaluate
situation and have a response to it. Fuller moves away from risk and task demands and difficulty,
respond to it. Threshold, at a certain moment there will be a response. People who lack emotions
should be irrational then but can perform well in tests!
Allostasis vs Homeostasis -> dynamic (changing motivational influences) vs fixed target level of task
difficulty or risk. Motivation has a influence in allostasis.
Capability and Task Demands = central element in RAT!!!
RAT has testable hypothesis -> evaluate task difficulty, risk, loss control by speed. Threshold is passed
eventually, more risk after that. Not linear relationship.
Zero-risk theory -> hardly experience risk, when risk then it is a warning. Threshold is the main thing,
there are margins. It is a learning theory. Criticized because risk is not continuously monitored so
how know that threshold is past. How the extra motives function is vague. Not easily tested! Need to
learn and create safety margin, but requires too many contingencies.
Risk/threat avoidance theories -> Risk is aversive and avoided. Behavioural contingencies are learnt.
What is risky depends on experience.
*RISK -> the probability of an outcome * the consequence of that outcome. Risk is not the same for
everybody. We are at risk but do not experience it every time!! The calculation requires the
knowledge level; effortful and biased.
Risk monitoring -> target level of risk; target feeling of risk/task difficulty or zero risk.
Safety Margin Model -> threshold model, motivated to be comfortable (threshold). It is time based:
keep constant, time to collision and line crossing (Summala). It is very unconscious, do automatically
and not think about it. Safety margins are monitored; personal space.
2 categories of perception:
-continuous/constant perception = RHT, Task Difficulty Homeostasis, RAT
-threshold perception (threshold past and then aware) = Zero Risk Theory, Threat avoidance theory
and multiple comfort zone model (Summala)
Michon’s Model (Janssen’s) => performance and decision making.
Rasmussen – human performance. We perform on different levels, skills (automated, no effort),
rules (looking for similarities and copy appropriate behaviour, if and then, pattern matching and
recognition, more effort and time) and knowledge (bounded rationality and thinking, don’t do it very
often, it is slow and takes effort and time).
, Behaviour starts at a high level but we want to operate as a skill level. By automating it, it becomes a
lower level and then it becomes easier and manageable. This in an energy efficient way of dealing
with the world.
Hierarchical decision making -> strategic level (long time, general plans), manoeuvring level
(seconds, controlled) and control level (automatic behaviour).
Research: measures on these different levels. Low/Control level -> Longitudinal control(speed),
lateral control (in lanes) and reaction time. SDLP = standard deviation of lateral position -> when to
much than problems, want a low SDLP, when it rises then more risk.
Manoeuvre/intermediate -> car following and gap acceptance ; when more impulsive behaviour
(XTC), then accept smaller gaps, they are willing to take more risk.
Strategic/top level -> route choice or mode of transport.
Conclusion: Skills are necessary but only to a point. Change behaviour, not attitudes (easier and more
effective). Do not expect rationally – people are not calculators and emotions play a role. A lot of
behaviour is habitual/well learnt (change is early). There are critical decision points (threshold
crossed). The best way to change behaviour is to change the environment in which people operate,
not to try and change the person directly.
