Traffic Psychology
LECTURE 1
- Testing models of driver behavior
Criticism of older models:
- Lack of testable hypothesis + over-reliance on risk
→ Michon (1985, 1989) preferred rule-based models
Ranney (2002) information-processing theories, but they lack of impact
→ Huguenin & Rumar (2001) criticized too broad or too specific
- task-related
- functional control
- motivational models
→ Rothengatter (2002) focused on attitudinal theories
+ criticized risk theories for: - circular reasoning
- emphasizing
- need for integrating social
psychology
Recent developments in DMR:
- New models aimed explaining behavioral adaptation
1. TDH (Task Difficulty Homeostasis theory)
RAT (Risk Allostasis Theory)
2. RMM (Risk Monitor Model)
3. Multiple Comfort Zone Model
1. TDH (Task DIfficulty Homeostasis Theory)
● Explains drivers prefer operating with range of task difficulty
- capability
- effort motivation
- trip goals
● Task-Capability Interface (TCI)
- Capability , A combination of physiological traits (reaction time),
learned experiences & human factors (motivation, fatigue)
- Task Demands , External conditions like road environment,
speed & other drivers’ behavior
- Task difficulty, arises when task demands approach or exceed
capability → leading to potential loss of control
RAT (Risk Allostasis Theory)
● expands on TDH by incorporating feelings of risk as a continuously monitored
variable that informs driver decisions.
, → SMH (Somatic Marker Hypothesis, underpins RAT suggesting unconscious
emotional body states guide rational decisions
● Criticisms
- RAT relies heavily on subjective perceptions of
risk and task difficulty, may not be continuously
monitored as suggested
- Comparisons to RHT (Risk Homeostasis
Theory) were made, but Fuller differentiates
RAT as focusing on feelings of risk.
2. RMM (Risk Monitor Model)
● RMM integrates components from previous models and emphasizes the role
of somatic markers in decision-making
→ it proposes that feelings drive behavior, aiming for a target ‘best feeling’ and
adapting behavior accordingly
● Evolutionary perspective → RMM views driving as part of survival
mechanism, monitoring risk through physiological responses
● Criticisms
- RMM may overstate the bodys role as a risk monitor and lacks
sufficient feedback loops for complex task like driving
3. Multiple Comfort Zone Model
● builds on zero-risk theory + focus on maintaining safety margins
→ safety margins act as a buffer for driver behavior & when breached drivers
experience discomfort → acts as a warning
● Comfort zones → feeling comfortable → not actively monitoring risk, unless
approaching limits of safety margins
● Satisficing → Making decisions by doing just enough to achieve goals, rather
than optimizing actions
● Criticisms
- Challenges from studies showing feelings of risk may be
constantly present rather than only activated of thresholds
! Move away from older trait-based models + unfalsifiable theories
! These modern models attempt overcoming shortcomings but still face challenges
, LECTURE 2
- Automated driving: Safety blind spots
● 2 main classes:
- Advanced Driver Assistance Systems (ADAS): Assists driving tasks
- Infotainment Systems: Supports auxiliary tasks like communication
→ 5 levels of automation 0=no automation, 5 = full automation
→ fully automated vehicles require: - public trust
- user acceptance
- regulatory frameworks
- extensive testing
● Human error is primary cause of road accidents
→ AD claims automation reduces accidents by eliminating human errors
→ But this claim assumes all crashes are due to human misjudgment &
that technology is error-free
→ Many crashes because of poor vehicle / road design
→ AD systems don't prevent all crashes if external factors no addressed
● Studies have been inconclusive regarding safety advantages
→ AV limited exposure + ideal testing conditions = difficult to generalize crash rates
- Computers are fast & precise, but decisions ≠ flawless
- Software failures lead to crashes (Google’s self-driving car & Tesla)
- AD may fail predicting human behavior or face challenges in
unforeseen situations
● Role of driver in AD
→ SAE levels are inadequate for assessing human-automation interaction
More practical categorization considers:
- If automation is partial or complete
- If it operates some of the time or all of the time
→ This framework clarifies ongoing role of driver, especially in systems partially auto.
● Can paradoxically increase driver workload in critical situations:
- Task Allocation → Left with more complex tasks, difficult to automate
- Deskilling → Reliance on automation can erode essential driving skills
- Cognition → May struggle with situational awareness when supervising
automated systems
- Control → Changes in vehicle dynamics make it harder to maintain control
- Lack of trust → May over-trust or under-trust AD, leading to misuse
- Liability → Determining responsibility for crashes involving AV remains
unclear
LECTURE 1
- Testing models of driver behavior
Criticism of older models:
- Lack of testable hypothesis + over-reliance on risk
→ Michon (1985, 1989) preferred rule-based models
Ranney (2002) information-processing theories, but they lack of impact
→ Huguenin & Rumar (2001) criticized too broad or too specific
- task-related
- functional control
- motivational models
→ Rothengatter (2002) focused on attitudinal theories
+ criticized risk theories for: - circular reasoning
- emphasizing
- need for integrating social
psychology
Recent developments in DMR:
- New models aimed explaining behavioral adaptation
1. TDH (Task Difficulty Homeostasis theory)
RAT (Risk Allostasis Theory)
2. RMM (Risk Monitor Model)
3. Multiple Comfort Zone Model
1. TDH (Task DIfficulty Homeostasis Theory)
● Explains drivers prefer operating with range of task difficulty
- capability
- effort motivation
- trip goals
● Task-Capability Interface (TCI)
- Capability , A combination of physiological traits (reaction time),
learned experiences & human factors (motivation, fatigue)
- Task Demands , External conditions like road environment,
speed & other drivers’ behavior
- Task difficulty, arises when task demands approach or exceed
capability → leading to potential loss of control
RAT (Risk Allostasis Theory)
● expands on TDH by incorporating feelings of risk as a continuously monitored
variable that informs driver decisions.
, → SMH (Somatic Marker Hypothesis, underpins RAT suggesting unconscious
emotional body states guide rational decisions
● Criticisms
- RAT relies heavily on subjective perceptions of
risk and task difficulty, may not be continuously
monitored as suggested
- Comparisons to RHT (Risk Homeostasis
Theory) were made, but Fuller differentiates
RAT as focusing on feelings of risk.
2. RMM (Risk Monitor Model)
● RMM integrates components from previous models and emphasizes the role
of somatic markers in decision-making
→ it proposes that feelings drive behavior, aiming for a target ‘best feeling’ and
adapting behavior accordingly
● Evolutionary perspective → RMM views driving as part of survival
mechanism, monitoring risk through physiological responses
● Criticisms
- RMM may overstate the bodys role as a risk monitor and lacks
sufficient feedback loops for complex task like driving
3. Multiple Comfort Zone Model
● builds on zero-risk theory + focus on maintaining safety margins
→ safety margins act as a buffer for driver behavior & when breached drivers
experience discomfort → acts as a warning
● Comfort zones → feeling comfortable → not actively monitoring risk, unless
approaching limits of safety margins
● Satisficing → Making decisions by doing just enough to achieve goals, rather
than optimizing actions
● Criticisms
- Challenges from studies showing feelings of risk may be
constantly present rather than only activated of thresholds
! Move away from older trait-based models + unfalsifiable theories
! These modern models attempt overcoming shortcomings but still face challenges
, LECTURE 2
- Automated driving: Safety blind spots
● 2 main classes:
- Advanced Driver Assistance Systems (ADAS): Assists driving tasks
- Infotainment Systems: Supports auxiliary tasks like communication
→ 5 levels of automation 0=no automation, 5 = full automation
→ fully automated vehicles require: - public trust
- user acceptance
- regulatory frameworks
- extensive testing
● Human error is primary cause of road accidents
→ AD claims automation reduces accidents by eliminating human errors
→ But this claim assumes all crashes are due to human misjudgment &
that technology is error-free
→ Many crashes because of poor vehicle / road design
→ AD systems don't prevent all crashes if external factors no addressed
● Studies have been inconclusive regarding safety advantages
→ AV limited exposure + ideal testing conditions = difficult to generalize crash rates
- Computers are fast & precise, but decisions ≠ flawless
- Software failures lead to crashes (Google’s self-driving car & Tesla)
- AD may fail predicting human behavior or face challenges in
unforeseen situations
● Role of driver in AD
→ SAE levels are inadequate for assessing human-automation interaction
More practical categorization considers:
- If automation is partial or complete
- If it operates some of the time or all of the time
→ This framework clarifies ongoing role of driver, especially in systems partially auto.
● Can paradoxically increase driver workload in critical situations:
- Task Allocation → Left with more complex tasks, difficult to automate
- Deskilling → Reliance on automation can erode essential driving skills
- Cognition → May struggle with situational awareness when supervising
automated systems
- Control → Changes in vehicle dynamics make it harder to maintain control
- Lack of trust → May over-trust or under-trust AD, leading to misuse
- Liability → Determining responsibility for crashes involving AV remains
unclear
