tom

CVEN4405: Human Factors in Civil and Transport Engineering

Human Performance Vulnerabilities and Traffic Engineering:

Driver distraction and Fatigue Term 3 2020

Week 8, Lecture 1a

1

CVEN 4405 Human Factors in Civil and Transport Engineering

Term 3 2020 Week 8 - Lecture 1a

PRESS RECORD BUTTON

+

SHARE SCREEN

2

Lecture Recordings

PLEASE NOTE.

All lectures today are being recorded.

Participation in this meeting indicates your consent to be included in the meeting recording.

3

Welcome Back!

4

Course Coordinator and Lecturer

Prof. Michael Regan, PhD Professor of Human Factors

Research Centre for Integrated Transport Innovation (rCITI) School of Civil and Environmental Engineering

University of NSW Sydney

T: +61 (0)2 9385 9504 E: [email protected]

Staff Webpage

5

The CVEN 4405 Teaching Team

Coordinator and Lecturer Prof. Michael Regan Professor of Human Factors Research Centre for Integrated Transport, UNSW Sydney E: [email protected]

Teaching Fellow Dr Prasannah Prabhakharan Research Fellow Research Centre for Integrated Transport, UNSW Sydney E: [email protected]

Demonstrator Mitch Cunningham E:[email protected]. au

6

Review of last lecture

• Delineation • Behavioural adaptation

7

This lecture - Overview

• Driver Distraction and Road Engineering • Driver Fatigue and Road Engineering

8

Learning Outcomes

CLO1: Explain the fundamental principles of HF that can be used by civil and transport engineers to facilitate user- centred design

CLO2: Apply HF principles, methods and data to the design of road and traffic management systems

CLO3: Plan for the integration of HF into the design lifecycle of the road and traffic management system

Driver Distraction and Road Engineering

10

Distraction and Inattention - Revision

• In Wickens’ model, attention is required for several stages of information processing - perception, working memory, decision making and response execution.

• It follows that, if we don’t pay attention to what we are doing (e.g. driving) – that is, if we are inattentive - some stages of information processing will not proceed properly, or at all, resulting in degraded performance and error.

•

11

Inattention – Revision (2)

• There are several different mechanisms of inattention.

• These mechanisms have been defined by Regan, Hallett & Gordon, 2011 (next slide)

• Distraction is just one mechanism of inattention…….

12

Model of Driver Inattention

Source: Regan, Hallet & Gordon, 2011

• Task-related thoughts

• Task-unrelated thoughts: o Internal/Intentional o Internal/ Unintentional o External/Intentional o External/Unintentional

• Daydreams

Internal competing activities

Driver Inattention

Driver Restricted Attention

Driver Misprioritised Attention

Driver Neglected Attention

Driver Cursory Attention

Driver Diverted Attention (Distraction)

Non-Driving-Related Driving-Related

13

Driver Distraction – Definition

Driver distraction has been defined as the:

• “Diversion of attention away from activities critical for safe driving toward a competing activity, which may result in insufficient or no attention to activities critical for safe driving”

Source: Regan, Hallett, & Gordon 2011, p. 1776

14

When a driver is distracted, they take their:

 Eyes off the road (“visual distraction”)  Mind (attention) off the road (“cognitive distraction”)  Ears off the road (“auditory distraction” – as a result of cognitive

distraction)  Hand(s) off the steering wheel (“physical interference”)

Driver Distraction – Mechanisms (2)

15

Driver Distraction – Impact on Performance (3) • As a result, driving performance is usually degraded unless the driver is

capable of timesharing between the driving task and the distracting task that is competing for their attention.

• If a driver is not capable of timesharing between the driving task and the distracting task that is competing for their attention, performance will be degraded and crash risk increases.

• It is estimated that distraction contributes to around 16% of serious injury and fatal crashes in Australia (Beanland et al., 2013)

• The next slides show some of the impacts of distraction on driving performance when drivers take their eyes and mind off the road.

16

Eyes off the Road: Impact on Driving

Selecting information: • miss relevant information from roadway • gaze concentration - when eyes return to the roadway

Processing information: • Longer reaction times to roadway warnings and braking lead

vehicles • change blindness – disrupts detection of changes in the

road

Source: Bayley et al. (2009); Horberry & Edquist (2009); Bruyas (2013); Victor et al (2009)

17

Eyes off the Road: Impact on Driving (2)

Driving performance: • lane keeping – worse; drive where you look • speed – reduce speed more with visual-manual distraction • following distance – increases more with visual-manual

distraction • more collisions

Source: Bayley et al. (2009); Horberry & Edquist (2009); Bruyas (2013); Victor et al (2009)

18

Mind off the Road: Impact on Driving

Selecting information: • gaze concentration – increased eye focus on road straight ahead • Less attention to peripheral field- eg neglect checking rear-view mirrors,

speedometer, peripheral hazards

Processing information: • “inattention blindness” – pedestrian walks from behind parked car; you

look at him, but don’t respond, or respond late • memory loss - fail to remember what has been seen during drive

Source: Bayley et al. (2009); Horberry & Edquist (2009); Bruyas (2013)

19

Mind off the Road: Impact on Driving (2)

Driving performance: • response time increases – resulting in more hard braking • look less at traffic lights – and miss red lights • more navigation errors • improved lane keeping performance – gaze concentration • no appreciable impact on following distances • accept shorter gaps when turning right across traffic • small decreases in speed; greater decreases if holding the phone

Source: Bayley et al. (2009); Horberry & Edquist (2009); Bruyas (2013)

20

Mind off the Road: Impact on Driving (3)

Driving performance (cont): • fewer lane changes – compensation • conflicts with vulnerable road users • more violations (speeding; red light running; crossing solid lines) • reduced ability to cope with wind gusts • errors – stopping at green lights; taking off before light green • Reduced scanning of intersection areas to the right • reduction in situation awareness – less able to identify, locate and

respond to hazardous vehicles and to avoid accidents.

Source: Bayley et al. (2009); Horberry & Edquist (2009); Bruyas (2013)

21

Hands off Wheel: Impact on Driving

Hypotheses: • one hand on wheel (automatic vehicle) = reduced steering

control during turns and emergency situations

• one hand on wheel (manual vehicle) = as above + reduced gear-changing control

• Steering with knees = as above + reduced speed control + longer RT to brake

Source: Regan, 2020

22

Objects (e.g., mobile phone; advertising billboard)

Events (eg lightning; explosion)

Passengers

Other road users and vehicles

Animals

Internal stimuli (that stimulate internal thought)

Source: Regan, Hallett & Gordon, 2011

Driver Distraction – All Sources

23

Driver distraction – Sources Outside the Vehicle (1) • Around 30% of distraction-related crashes derive from

driver engagement with sources of distraction that are from outside the vehicle

(Gordon, 2009).

• These sources of distraction include: (next slide)

24

Driver distraction – External Sources (2)

Source: Gordon, 2009)

• Animals • Architecture • Advertising signage • Construction zones/equipment • Crash scenes • Incidents (e.g. road rage) • Insects • Landmarks • road signs • road users • Scenery • Other vehicles • weather (e.g. lightning).

25

Driver distraction (3)

• Road and traffic engineers have little or no control over sources of distraction within the vehicle

• They do, however, have some scope to manage distraction from some sources of distraction residing outside the vehicle

• “Traffic engineering has a role in preventing serious crashes arising from driver distraction and fatigue”

Source: PIARC, 2015

26

Distraction - PIARC Road Design Recommendations

1. Lower energies through conflict points to within human tolerances

i.e. in the event of driver distraction (or fatigue), infrastructure measures will generally ensure that vehicle speeds are within the human tolerances for serious injury through relevant conflict points

Source: PIARC, 2015

27

PIARC Recommendations (2)

2. Design to provide opportunities for road users to recover from mistakes and non-compliance.

• i.e. provide opportunities for crashes to be avoided in the case of driver distraction (or fatigue)

• e.g. locating crash barriers further from the through traffic lanes provides an opportunity for errant vehicles to recover before hitting the barrier

Source: PIARC, 2015

28

PIARC Recommendations (3)

• 3. Design to lower the risk of a crash occurring to an “acceptable” level

• i.e. Design the road to minimise the risk of driver distraction (or fatigue) from occurring in the first place.

• e.g. by paying careful attention to human factors - such as preventing the road from surprising the road user (e.g. with a concealed driveway)

Source: PIARC, 2015

29

PIARC Recommendations (4)

Human tolerances for serious injury crashes. There is a 10% likelihood of a fatality at the following travel speeds:

Source: PIARC, 2015

Type of Infrastructure Travel Speed (km/hr) Locations with possible conflicts between pedestrians and cars

30

Intersections with possible side impacts between cars

50

Roads with possible frontal impacts between cars 70

Roads with no possibility of a side impact or frontal impact (only impact with the infrastructure)

100 +

30

PIARC Recommendations – Specific Treatments PIARC (2016) recommends a range of road engineering treatments that can be used to mitigate the effects of driver distraction (and fatigue):

Hierarchy Level 1 treatments: concrete side barriers; steel side barriers; wire rope side and median barriers; lateral shift of road; roundabouts; grade separation at intersections; speed humps;

Source: PIARC, 2015

31

PIARC Recommendations (2)

Hierarchy Level 2 treatments: rumble strips; tactile line markings; speed humps; rough shoulders; variable speed signs

Source: PIARC, 2015

32

Driver distraction – Advertising Billboards

• Advertising billboards are one of the external sources of distraction described earlier.

• Advertising billboards, especially dynamic billboards with moving features, have potential to attract more, and longer, eye glances off the forward roadway than other typical traffic signs (Dukic et al., 2013)

• Dynamic advertising billboards are more likely than static billboards to attract long eye glances i.e. > 2 s (Decker et al., 2014)

33

Driver distraction – Advertising Billboards (2)

Eye glances off the forward roadway that are > 2 s have been shown to double crash risk !!

Source: Klauer et al., 2006

34

Design and content of roadside advertising – Design Guidance Guidance for the design and content of roadside advertising (Roberts et al., 2013):

• Advertising should not contain movement or flashing features – as these are particularly good at capturing driver attention involuntarily

• Advertising should not contain emotional content – as this is effective in capturing attention and creating cognitive distraction (‘mind off road’), even for some time after the advertising has been passed (e.g. Chan & Singal, 2013)

35

Design and content of roadside advertising – Design Guidance (2)

• Advertising should not contain too much textual information – too much text will encourage longer driver eyes-off-road time

• The amount of text should be easily read by an approaching driver within a 2 second time period.

36

Location and placement of roadside advertising – Design guidance Guidance for the location and placement of roadside advertising (Roberts et al., 2013):

• advertising devices should not be located in such a way that they interfere with the effectiveness of a traffic control device (e.g. by restricting sightlines or distracting from traffic control devices via proximity or as a background).

• advertising devices should not be located so that they are visible at the approach to, or from, an intersection, pedestrian crossing, tram stop or in any location that is likely to be highly demanding of attention.

Source: Roberts et al., 2013

37

Location and placement of roadside advertising – Design guidance (2)

Source: Roberts et al., 2013

• only one advertising device should be visible to drivers at any time.

• advertising devices should be placed so that enough time is available on approach for drivers to comprehend the message i.e. the sight distance must correspond to the required legibility distance.

38

Design guidance for other potential sources of driver distraction

• The PIARC (2016) document provides some general guidance for road design to prevent and mitigate the effects of distraction

• Cunningham, Regan & Cairney (2017) provide more specific guidance for some of the specific external sources of distraction listed earlier that traffic engineers may have some control over (next slides)

39

Design guidance for driver distraction - Animals

• On road sections where it is known that roadway incursions by animals are problematic, and could distract drivers:

• can use warning signs, and perhaps barriers, to minimise the likelihood of interactions between drivers and animals that distract drivers

Source: Cunningham, Regan & Cairney (2017)

40

Design guidance for driver distraction - Scenery

• Scenic routes and tourist roads are, by definition distracting, and are often located along winding rural roads.

• Traffic engineers should alert drivers to the potential for distraction along such roads.

• The most basic measure would be to reduce speed limits along scenic routes, to give drivers more time to:

• recover from the effects of distraction, and • reduce impact speeds in the event of a distraction-related

crash

What others can you think of?

Source: Cunningham, Regan & Cairney (2017)

41

Design guidance for driver distraction - Architecture • There are some grand buildings and monuments that have

potential to distract people.

• There isn’t much the traffic engineer can do about this.

• However, it may possible to visually mask (for example, with trees), prominent architectural structures and features that are known to distract drivers in particular locations

Source: Cunningham, Regan & Cairney (2017)

42

Design guidance for driver distraction - Crash scenes • “Rubber necking” is common driver behaviour around

crash scenes.

• It distracts drivers and is also known to cause crashes; and create traffic congestion downstream.

• Possible countermeasures include: • routing traffic away from crash scenes, where

possible • visually masking the scene in some way from

approaching traffic.

Source: Cunningham, Regan & Cairney (2017)

43

Design guidance for driver distraction - Road signs• Poorly designed traffic signs can, themselves, distract

drivers. • e.g. if they are absent in locations they should be

(e.g. no street name on the road you are turning onto), they may encourage drivers to adopt compensatory search strategies that distract them.

• e.g. if they are poorly designed – contain too much information; incomprehensible etc

• Poorly designed and absent road signs should be avoided.

Source: Cunningham, Regan & Cairney (2017)

44

Design guidance for driver distraction - Error tolerant road design

• Ultimately, the traffic engineer should strive for a distraction-tolerant road system such that, in the event of a distraction-related crash, no driver or other road user is killed or seriously injured (Tingvall, Eckstein, & Hammer, 2009).

• The PIARC (2015) document, discussed earlier, provides guidance for treatments that can achieve this.

Fatigue and Road Engineering

46

Fatigue

Fatigue is a factor that can impair the performance of any human activity.

We can distinguish between two types of fatigue: 1. Muscular fatigue 2. General fatigue

We talked about muscular fatigue earlier in the last lecture. It is a painful phenomenon that occurs in over-stressed muscles and is localised.

Source: Oborne (1987), p. 41

47

General Fatigue: Definition

General fatigue is, in contrast to Muscular fatigue, • “…a diffused sensation, which is accompanied by feelings of

indolence and disinclination for any kind of activity.”

• Discussion: Who has experienced general fatigue? • What are the symptoms?

Source: Fuller & Santos (2002), p. 87; Grandjean (1981), p. 168

48

General Fatigue: Causes

The major causes of generalised fatigue are: • Inadequate sleep • Inadequate rest or recovery • Illness

Source: Fuller & Santos (2002), p. 87

49

General Fatigue: Symptoms

The symptoms of general fatigue include: • Subjective feelings of weariness • Decreased motivation • Unwillingness to work • Slowed or impaired perception • Restricted field of attention • Decreased performance, in the form of irregularities in

timing and speed

» Discussion: What are some of the things that cause fatigue, in the rail environment?

Source: Fuller & Santos (2002), p. 87; Grandjean (1981), p. 168

50

Fatigue and Drowsiness

The words “drowsiness” and “fatigue” are often used interchangeably, but they are not the same.

While general fatigue can refer to a feeling of tiredness or exhaustion due to inadequate rest, prolonged work or illness, drowsiness refers specifically to the state just before sleep.

You can feel fatigued without feeling sleepy. However, lack of sleep does contribute to general fatigue.

Source: Fuller & Santos (2002), p. 87; https://www.optalert.com/drowsiness-vs-fatigue-how- do-they-differ/

51

General Fatigue: Time of Day

People are most alert: • In the mid-morning • In the early evening

People are least alert: • Between midnight a 5am • In the early afternoon (post lunch dip)

Source: PIARC (2016)

52

General Fatigue: Circadian Rhythm

Biological factors can have a substantial impact on fatigue, such as the natural circadian rhythm and the amount of sleep they have had.

Fatigue (and sleep-related accidents) most likely to occur in the early hours of the morning between 2am and 6am.

Source: Williamson et al., (2011); Horne & Reyner (1995); Pack et al., (1995); Eskandarian et al., (2007)

53

General Fatigue: Inadequate Sleep

People differ in the amount of sleep they need; the amount of sleep needed usually declines with age.

Sleep loss, especially if over several nights, can: • Seriously affect human performance

– E.g. memory, decision making, reaction time, mood

• Be stressful, due to the effort of trying to stay awake • Cause people to fall asleep for a few seconds or less

(“microsleeps”). • Cause people to fall asleep completely

Source: Fuller & Santos (2002), p. 87

54

General Fatigue: Inadequate Sleep (2)

Microsleeps may not be noticed but can be dangerous when they occur during the performance of tasks that require continuous attention - like driving or monitoring tasks.

Sleep loss can cause drivers to fall asleep at the wheel. • The onset of sleep can happen very suddenly. • The period just before falling asleep is dangerous,

because drivers have no control over it.

Source: Fuller & Santos (2002), p. 87

55

General Fatigue: Inadequate Rest and Recovery

Rest and Recovery allows the body to: • replenish muscular energy stores • repair damaged tissues • removes waste from the brain • reduces inflammation

Without sufficient time to repair and replenish, the body cannot perform these tasks, leading to performance decrements.

Source: Kreher & Schwartz (2012), scientificamerica.com

56

General Fatigue: Effects of Illness

As we have all experienced, illness can also lead to general fatigue.

• E.g. it is a symptom of COVID-19

However, the focus is of this course is on design (e.g., job design, rostering etc) to minimise potential for fatigue.

57

Driver Fatigue: Road Safety

• Fatigue effects has been done very well researched, particularly in relation to road safety.

• It has been estimated that fatigue is a contributing factor in about 20-30% of road casualties in Australia.

• Findings from a recent ‘naturalistic driving study’ in the U.S. showed that driver fatigue/drowsiness increased crash risk by 3.4 times.

• Driver fatigue, like driver distraction, is a significant road safety issue.

Source: Australian Transport Council, (2011); Dingus et al., (2016); Cunningham, Regan & Cairney (2017)

58

Driver Fatigue: Causes

Driver fatigue may arise from: • driving at particular times of the day or night • driving for prolonged periods • driving in monotonous conditions

– e.g. open highways • driving in complex traffic environments

– e.g. bad weather; at night; unpredictable traffic

Source: Cunningham, Regan & Cairney (2017); PIARC (2016)

59

Driver Fatigue: Crash Characteristics

Fatigue-related road crashes typically have several defining characteristics:

• they occur late at night, early morning or mid-afternoon • they result in higher than expected crash severity • they involve a single vehicle leaving the roadway • They occur on a high speed road • the driver did not attempt to avoid the crash • the driver was the sole vehicle occupant

Source: US Expert Panel on Driver Fatigue and Sleepiness (1997); cited in PIARC (2016)

60

Driver Fatigue: PIARC Design Recommendations

The PIARC (2016) road design recommendations for fatigue are the same as those for distraction: i.e.

1. Lower energies through conflict points to within human tolerances (by setting speed tolerances)

2. Design to provide opportunities for road users to recover from mistakes and non-compliance.

3. Design to lower the risk of a crash occurring to an “acceptable” level

Source: PIARC (2016)

61

Driver Fatigue: Rest Areas

Rest areas have proven safety benefits.

They are located along or beside roadway to: • provide driver with an opportunity to stop driving and sleep/rest/stretch • help driver to counteract the effects of driver fatigue

Careful consideration needs to be given to: • the location of rest areas, particularly along high-speed roads • the best type of facilities to be provided at different locations

Source: PIARC (2016)

62

Driver Fatigue: Rest Areas (2)

PIARC (2016) recommendations: • Provide rest areas at regular intervals • Minor rest areas should be located at maximum intervals of

50 km • Major rest areas, with services, should be located at

maximum intervals of 100 km • Provide advance sings offering information relating to

upcoming distance and next rest areas

Source: PIARC (2016)

63

Driver Fatigue: Rest Areas (3)

Design Guidance Access into and egress from rest areas are important design

aspects for consideration:

• access and egress must be safe for vehicles entering or leaving the rest area and re-entering the traffic flow

• application and requirements will vary greatly from site to site

Source: Department of Transport and Main Roads (2014); cited in Cunningham, Regan & Cairney (2017)

64

Driver Fatigue: Monotony Reduction Treatments Monotony reduction treatments have been developed, but not

all have been evaluated: e.g.

• creating ‘sinuous, rhythmic road alignments’ – i.e. gently winding roads to counter monotony by providing a constantly changing visual field

• avoiding monotonous vegetation and roadside infrastructure along freeways

Source: PIARC (2016), p. 37

65

Driver Fatigue: Monotony Reduction Treatments (2)

Install simple objects, such as roadside art or signage, to increase mental stimulation and counteract fatigue along long and monotonous roadways

Must ensure, however, that such installations are located in areas that do not distract drivers from activities critical for safe driving.

Source: Roberts and Turner, (2008) cited in Cunningham, Regan & Cairney (2017)

66

Driver Fatigue: Signage and Road Markings

• Traffic signs and road markings can be used to:

• indicate that a stretch of road has a significant fatigue- related crash risk

• identify opportunities to stop (and rest) at designated rest areas or other locations (e.g., service station), with plenty of advanced notice.

Source: Cunningham, Regan & Cairney (2017)

67

Driver Fatigue: Rumble Strips

• Audio-tactile line markings (“rumble strips”) are small raised bumps on or adjacent to line markings. They:

• alert drivers through both sound and vibration when drifting out of their lane into oncoming traffic or off the edge of the road

• have been proven to be highly effective in reducing crashes

• Can be difficult for them to generate a sufficiently loud signal to be effective for heavy vehicles (e.g. trucks)

Source: Wooley & McLean (2006); Cunningham, Regan & Cairney (2017)

68

Road and traffic management design for other human performance vulnerabilities • We talked earlier in the course about other some human

performance vulnerabilities: emotion, arousal, stress, mental workload, alcohol, and drugs

• Various design recommendations discussed during the course will ensure that the road and traffic environment does not stress the driver or mentally overload them.

Questions ?

Over to ….

Prof. Michael Regan, PhD Research Centre for Integrated Transport Innovation

(rCITI) Room 112, Civil Engineering Building (H20)

E: [email protected]

CVEN4405: Human Factors in Civil and Transport Engineering

Human Factors Research: Data Collection Methods

Week 8, Lecture 1b

72

CVEN 4405 Human Factors in Civil and Transport Engineering

Term 3 2020 Week 8 - Lecture 1b

PRESS RECORD BUTTON

+

SHARE SCREEN

73

Lecture Recordings

PLEASE NOTE.

All lectures today are being recorded.

Participation in this meeting indicates your consent to be included in the meeting recording.

74

Welcome Back!

75

Course Coordinator and Lecturer

Prof. Michael Regan, PhD Professor of Human Factors

Research Centre for Integrated Transport Innovation (rCITI) School of Civil and Environmental Engineering

University of NSW Sydney

T: +61 (0)2 9385 9504 E: [email protected]

Staff Webpage

76

The CVEN 4405 Teaching Team

Coordinator and Lecturer Prof. Michael Regan Professor of Human Factors Research Centre for Integrated Transport, UNSW Sydney E: [email protected]

Teaching Fellow Dr Prasannah Prabhakharan Research Fellow Research Centre for Integrated Transport, UNSW Sydney E: [email protected]

Demonstrator Mitch Cunningham E:[email protected]. au

77

Review of last lecture

• Driver Distraction and Road Engineering • Driver Fatigue and Road Engineering

78

This lecture - Overview

• Human Factors data sources • Data Collection Methods:

• Descriptive studies • Experimental studies • Evaluation studies (next lecture)

79

Learning Outcomes

CLO1: Explain the fundamental principles of HF that can be used by civil and transport engineers to facilitate user- centred design

CLO2: Apply HF principles, methods and data to the design of road and traffic management systems

CLO3: Plan for the integration of HF into the design lifecycle of the road and traffic management system

Human Factors Data Sources

81

Human Factors Data Sources

The Human Factors discipline applies data about human capabilities and limitations to the design of objects, products, environments and systems used by people

Human Factors specialists rely on a number of sources of information to guide their involvement in the design process.

Source: Sanders and McCormick (1987); Wickens et al, 2004

82

Human Factors Data Sources (2)

These include: Data Compendiums: • e.g. containing tables and formulae about human capabilities and

limitations

Human Factors Design Standards: • contain detailed requirements for human-centred design: controls,

displays, labelling, anthropometry, workspace design, environmental factors, maintenance, hazards and safety.

Source: Wickens et al (2004), p. 36

83

Intro. to Human Factors Research Methods

Human Factors Principles and Guidelines: • These exist for a variety of topics:

– products (e.g. Norman, 1992) – equipment design (e.g. Van Cott & Kinkade, 1972) – physical facilities (e.g. McVey, 1990) – visual display units (e.g. Gilmore, 1985) – software interfaces (e.g. Schneiderman, 1992) – information system in cars (e.g. Campbell et al., 1999). – Human Factors Guidelines for Road Systems (Campbell, 2012)

Source: Cited in Wickens et al (2004), p. 36

84

Human Factors Guidelines for Road Systems

Source: https://www.nap.edu/catalog/22706/human-factors-guidelines-for-road-systems- second-edition

85

Intro. to Human Factors Research Methods (4)

Human factors specialists also do their own research to: • collect data to inform the user-centred design process • evaluate the quality of their designs, or the quality of things

designed by others

Hence, research is important in informing user-centred design and in evaluating designs.

Source: Sanders and McCormick (1987), p. 20

Data Collection Methods

87

Human Factors Research - Classification

Human Factors research can be classified into 3 types:

1. Descriptive studies

2. Experimental studies

3. Evaluation studies (in Lecture 4b)

Source: Sanders and McCormick (1987), p. 21

88

HF Research – Choosing the Right Type

Choosing which of the three methods to use depends on several factors:

Source: Sanders and McCormick (1987), p. 21

Effectiveness whether the method accomplishes its purpose Ease of use how easy the methodology is to use Flexibility how flexibility it can be used in different situations/contexts Range the no. of phenomena, behaviours, events etc that can be

measured Validity the degree to which data produced are like those that occur in

real life Reliability the degree to which the data produced are consistent over time

and between applications Objectivity the extent to which the method relies on data/procedures external

to the person who is applying the method

Descriptive Studies

90

Descriptive Studies

Descriptive studies aim to characterise a population (in our case, people) in terms of certain attributes; that is, to describe a particular population of people

Descriptive studies can answer what, when, where, when and how questions, but not why questions.

Descriptive studies provides basic data upon which many design decisions are made.

Source: Sanders and McCormick (1987), p. 22; https://www.scribbr.com/methodology/descriptive-research/

91

Descriptive Studies: Examples of Descriptive Research Questions • What are the attributes of people who cross pedestrian

crossings illegally? • How old are people in the Australian driving population? • What are the most popular social media sites for under-

18s? • How prevalent is speeding in the Australian population? • What are the most significant risks that contribute to road

crashes in Australia?

Source: https://www.scribbr.com/methodology/descriptive-research/

92

Descriptive Studies: Collecting Data

Data for descriptive studies can be collected: • in the field (e.g. characterising the attributes of pedestrians

who cross pedestrian crossings illegally)

• In the laboratory (e.g. an anthropometric study of body dimensions)

Source: Sanders and McCormick (1987)

93

Descriptive Studies: Collecting Data (2)

Methods used to collect descriptive data include: • Surveys and questionnaires: for large volumes of self-reported data;

relies on honesty and accuracy of respondents

• Observations: for actual data on behaviours, events, phenomena etc; don’t have to rely on honesty and accuracy of participants

• Case studies: for detailed data on a narrowly-defined subject.

• Interviews/focus groups: for detailed data on a narrowly-defined subject.

Source: https://www.scribbr.com/methodology/descriptive-research/

94

Descriptive Studies: Choosing a Research Setting • Usually, descriptive studies are done in the real world rather

than in a laboratory. • e.g. a study in which we observe number of pedestrians who

walk through red lights illegally and characterise them by: – Age – Gender – Race

Source: Sanders and McCormick (1987), p. 22

95

Descriptive Studies: Selecting Variables

In descriptive studies, we measure 2 types of variables: 1. Criterion variables (which describe physical

characteristics and behaviours) – Physical characteristics – e.g. weight – Performance data – e.g. reaction time; visual acuity – Subjective data – e.g. preferences, opinions, ratings – Physiological data – e.g. heart rate, body temperature

2. Stratification variables (normally collected in surveys)

– E.g. age, sex, level of education

Source: Sanders and McCormick (1987), p. 22

96

Descriptive Studies: Selecting Study Participants Choosing a proper sample for descriptive studies is critical, to

ensure that the findings from the study are valid.

There are 3 main considerations: 1. Representative sample

2. Random sampling

3. Sample size

Source: Sanders and McCormick (1987), p. 23

97

Descriptive Studies: Selecting Study Participants (2) 1. Representative sample • you need to collect data from a sample of people representative of the

population of interest

• It is representative if it contains all aspects of the population in the same proportion as the real population. – E.g. If, in population of pedestrians in NSW, 30% are under 21 yrs,

40% are between 21 and 40, and 40% are over 40 yrs, then the sample needs to contain the same % for each group

• A sample that is not representative, is biased.

Source: Sanders and McCormick (1987), p. 23

98

Descriptive Studies: Selecting Study Participants (3) 2. Random sampling • In order to obtain a representative sample, it is necessary to select the

sample randomly from the population.

• A random sample is selected when each member of the population has an equal chance of being included in the sample.

• Although in the real world it is difficult or impossible to select a truly random sample.

Source: Sanders and McCormick (1987), p. 23

99

Descriptive Studies: Selecting Study Participants (4) 2. Random sampling (cont.) • e.g. We want to select a random sample of 200 students from UNSW to

participate in an anthropometric survey: – Need to identify all students in the university population (assume

10,000) – Assign each of them a number from 1 to 10,000 – Use a random number generator to select 200 random numbers – Select those numbers from the 10,000 numbers representing the

student population. – Now you have your random sample!

Source: Sanders and McCormick (1987), p. 23

100

Descriptive Studies: Selecting Study Participants (5) 3. Sample size • Determining how many people to sample for a descriptive

study is also critical. – Too small = may include a disproportionate number of individuals

which are outliers and anomalies. These skew the results and you don’t get a fair picture of the whole population.

– Too big = too complex, expensive and time-consuming to run, and although the results are more accurate, the benefits don’t outweigh the costs.

Handy website to calculate sample size, (when population size is known): https://www.qualtrics.com/au/experience- management/research/determine-sample-size/

Source: Sanders and McCormick (1987), p. 23

101

Descriptive Studies: Selecting Study Participants (6) 3. Sample size (cont.) The number of people required for the sample depends on 3 things:

i. The degree of accuracy required of the data – the greater the accuracy required, the bigger the sample needs to be.

ii. Variance in the population – the greater the degree of variability in the measure that is being collected (e.g. body height), the larger the sample size needs to be.

iii. The statistic being estimated (e.g. mean; 50th percentile) – e.g. more people are needed to estimate the 50th percentile (median) height for a population than for the mean.

Source: Sanders and McCormick (1987), p. 23

102

Descriptive Studies: Analysis of Data

For descriptive studies, only basic statistics are normally computed.

The following are the most common: • Means (averages) • Standard Deviations • Correlations • Percentiles

Source: Sanders and McCormick (1987), p. 24

103

Means and Standard Deviation

Mean is the average value • the sum of n values divided by n.

The standard deviation (SD) a measure of the variability of a set of numbers around the mean. e.g.

• we measure the reaction times to an unexpected road hazard for a group of drivers

• the reaction times vary across drivers

If they vary greatly, the standard deviation is large; if they are close together, the SD would be small

Source: Sanders and McCormick (1987), p. 24

104

Means and Standard Deviation

In a normal distribution (e.g. of reaction times): • Around 68% of RTs fall within +/- 1SD of the mean • Around 95% of RTs would fall within +/- 2SD of the mean • Around 99.7% of RTs would fall within +/- 3SD of the mean

Source: Sanders and McCormick, 1987, p. 24; Google Images

105

Correlation

A correlation is a measure of the degree of relationship between 2 variables.

A correlation co-efficient (e.g. Pearson’s product-moment correlation, r) is computed, which indicates the degree to which the variables are related linearly.

Correlations range from +1.0 (perfect positive relationship) to - 1.0 (perfect negative relationship).

Source: Sanders and McCormick (1987), p. 24

106

Correlation: Positive Relationship

A positive relationship means high values on one variable are associated with high values on the other variable

• e.g. height and weight; education and income

Source: Sanders and McCormick (1987), p. 25

107

Correlation: Negative Relationship

A negative relationship means high values on one variable are associated with low values on the other variable

• e.g. age and strength

Source: Sanders and McCormick (1987), p. 25

108

Correlation: No Relationship

Two variables can also have no relationship • high values on one variable are not associated, in any

meaningful way, with the other variable

109

Correlation does not imply causation!

Just because two variables are correlated, does not mean they are causally related to each other.

Source: https://www.tylervigen.com/spurious-correlations

110

Percentiles

“Percentiles correspond to the value of a variable below which a specific percentage of the group fall”

• e.g. if the 5th percentile standing height for males is 162 cm, this means that only 5% of males are shorted than 162 cm.

The 50th percentile height is the same as the median – the height at which 50% of males are taller and 50% are shorted.

The interquartile range is the range from the 25th to 75th percentiles.

Percentiles are important in using anthropometric data

Source: Sanders and McCormick (1987), p. 25

Experimental Studies

112

Experimental Studies

“The purpose of experimental research is test the effects of some variable on human behaviour.”

• e.g. the effect of alcohol on lane keeping behaviour • e.g. the effect of speed humps on speeding behaviour • e.g. the effect of delineation on lane keeping behaviour around curves • e.g. the effect of stop signs on stopping behaviour

Experimental studies are usually concerned with assessing the effect of a variable on human behaviours - unlike descriptive studies, which are concerned with describing population parameters and behaviours.

Source: Sanders and McCormick (1987), p. 26

113

Experimental Studies: Research Setting

As with descriptive studies, experimental studies can be conducted:

• In the field (e.g. the effects of speed humps on speeding)

• In the laboratory (e.g. the effects of speed humps on speeding)

Source: Sanders and McCormick (1987), p. 26

114

Experimental Studies: Research Setting (2)

Field experiments (e.g. the effects of speed humps on speeding) are more realistic in terms of:

• the relevant task variable examined (e.g. speed hump)

• the behaviour being studied (e.g. speed control)

• environmental constraints (e.g. road surface condition)

• motivation of the drivers being studied (e.g. running late to work)

Source: Sanders and McCormick (1987), p. 26

115

Experimental Studies: Research Setting (3)

Laboratory experiments, on the other hand: • Give you experimental control over extraneous variables

(e.g. police enforcement) that could also affect the behaviour of interest (e.g. speed)

• The laboratory experiment can be replicated/repeated easily and quickly

Source: Sanders and McCormick (1987), p. 26

116

Experimental Studies: Research Setting (4)

Field experiments are usually: • much more costly than laboratory (e.g. simulator) experiments • might be unsafe for participants (e.g. if investigating the effect of alcohol

on lane keeping performance) • Lacking in experimental control (e.g. variables other than the speed

hump that might affect the speed behaviour of drivers, such as surrounding traffic, the presence of Police, bad weather, etc)

Source: Sanders and McCormick (1987), p. 26

117

Experimental Studies: Research Setting (5)

Laboratory experiments, however: • may lack realism • their findings may not be generalisable to the real world; although

modern virtual reality driving simulators look and feel increasingly realistic for experimental research

118

Experimental Studies: Selecting Variables

In experimental studies, the researcher: • manipulates one or more variables (e.g. alcohol)…

• in order to assess their effects on behaviours (e.g. lane keeping performance)…

• that are measured (e.g. in a driving simulator)…

• while other variables (e.g. that could also affect lane keeping performance) are controlled.

Source: Sanders and McCormick (1987), p. 26

119

Experimental Studies: Selecting Variables (2)

Independent variables (IVs) - The variables being manipulated (e.g. level of alcohol)

Dependent variables (DVs) - The variable being measured to assess the effects of the independent variable (e.g. no. of lane excursions)

Extraneous (or Confounding) Variables - The variables that are controlled for (e.g. drowsiness)

• Could also influence the DV (e.g. no. of lane excursions). • Controlled so that their effect is not confused (confounded) with the

effect of the IV Source: Sanders and McCormick (1987), p. 27

120

Experimental Studies: Choosing Participants

In experimental research, like descriptive research, the participants selected need to be representative of the people to whom the research findings will be generalised.

“The question in experimental studies is whether the subjects will be effected by the IV in the same way as the target population”

Source: Sanders and McCormick (1987), p. 27

121

Experimental Studies: Choosing Participants

If you wanted to conduct an experiment to assess the effect of road width on travel speed, you would have a wide choice of participants from which to choose.

• The characteristics of the drivers (e.g. age, gender) are not important in order to be representative.

However, assessing the effect of old age on the ability of drivers to comprehend messages on traffic signs would require a restrict choice of participants from which to choose

• You would need to choose a sample of older drivers only, in order to be representative.

122

Experimental Studies: Sample Size

For experimental research, the issue in determining sample size is to “collect enough data to reliably assess the effects of the independent variable with minimum cost in time and resources”

Generally, far fewer people are needed for experimental research than for descriptive research.

• E.g. In a typical simulator experiment, in which we assess the effect of alcohol (no alcohol versus 0.05 BAC) on lane keeping behaviour (e.g. no. of lane exceedances), 20 participants would be sufficient.

Source: Sanders and McCormick (1987), p. 27

123

Experimental Studies: How to Collect Data

Experimental data can be collected in the field or in a laboratory environment (usually a driving simulator for road safety research)

• E.g. in a driving simulator • E.g. In a instrumented real vehicles

124

Experimental Studies: Steps in Designing a Study 1. Define the Problem and Hypothesis: • e.g. research question: will widening a road increase existing travel

speeds, due to behavioural adaptation? • e.g. hypothesis: widening the road will increase existing travel speeds

due to behavioural adaptation

2. Specify Experimental Plan: i.e. specify details of experiment:

• the IV, the DVs • The experimental design • The experimental procedures (i.e. order of proceedings) • Tasks to be performed by our participants

125

Experimental Studies: Steps in Designing a Study 3. Conduct the Study:

» Obtain ethics approval » Develop study materials (e.g. questionnaires) » Develop study equipment (e.g. simulation scenario; eye trackers) » Recruit participants » Seek their consent » Pilot the experiment » Run study » De-brief participants

4. Analyse Data 5. Draw Conclusions – about the cause and effect

relationship

126

Experimental Studies: 3 Basic Designs

For any experiment, there three basic designs that can be used to collect data on cause and effect relationships:

1. Independent Group Design (or Between-Subjects Design)

2. Matched Pairs Design (or Matched Between-Subjects Design)

3. Repeated Measures Design (or Within Subjects Design)

127

Experimental Studies: 3 Basic Designs (2)

Source: https://www.youtube.com/watch?v=WnBOTsP8z4g

128

Experimental Studies: Analysing Data

Data from experimental studies are analysed using some type of inferential statistic

• e.g. chi square analysis ; t-tests; analysis of variance (ANOVA)

We won’t go into the actual statistics. There are resources you can refer to learn more about statistics.

The outcome of the statistical analysis is usually a statement about the statistical significance of the data.

Source: Sanders and McCormick (1987), p. 28

129

Experimental Studies: Analysing Data (2)

Statistical significance: • researchers often make statements like “the width of the

road had a significant effect on drivers’ speed choice”

• In other words, they are saying that the IV (lane width) had a significant effect on the DV (speed)

• Something is statistically significant if “there is a low probability that the observed effect, or difference between the means, was due to chance”

Source: Sanders and McCormick (1987), p. 28

130

Experimental Studies: Analysing Data (3)

Statistical significance (cont): • So if it is unlikely that the effect (increased speed) was due to chance, it

can be concluded that it was due to the effect of the IV (increased lane width).

• In statistics, we say that, if the results we obtained could have occurred only 5 times or less out of 100 by chance alone, then they are unlikely to be due to chance – and we conclude that the results are significant at the 0.05 level.

• If the results could have occurred more than 5 times out of 100 by chance, then we conclude that it is likely that chance was the cause of the effect on the DV, not the IV.

Source: Sanders and McCormick (1987), p. 28

131

Questions?

Over to ….

Prof. Michael Regan, PhD Research Centre for Integrated Transport Innovation

(rCITI) Room 112, Civil Engineering Building (H20)

E: [email protected]

CVEN4405: Human Factors in Civil and Transport Engineering

Human Factors Research: Evaluation Methods

Term 3, 2020

Week 8, Lecture 2

134

CVEN 4405 Human Factors in Civil and Transport Engineering

Term 3 2020 Week 8 - Lecture 2

PRESS RECORD BUTTON

+

SHARE SCREEN

135

Lecture Recordings

PLEASE NOTE.

All lectures today are being recorded.

Participation in this meeting indicates your consent to be included in the meeting recording.

136

Welcome Back!

137

Course Coordinator and Lecturer

Prof. Michael Regan, PhD Professor of Human Factors

Research Centre for Integrated Transport Innovation (rCITI) School of Civil and Environmental Engineering

University of NSW Sydney

T: +61 (0)2 9385 9504 E: [email protected]

Staff Webpage

138

The CVEN 4405 Teaching Team

Coordinator and Lecturer Prof. Michael Regan Professor of Human Factors Research Centre for Integrated Transport, UNSW Sydney E: [email protected]

Teaching Fellow Dr Prasannah Prabhakharan Research Fellow Research Centre for Integrated Transport, UNSW Sydney E: [email protected]

Demonstrator Mitch Cunningham E:[email protected]. au

139

Review of Last Lecture

• Human Factors data sources • Types of Human Factors research studies:

• Descriptive studies • Experimental studies • Evaluation studies (next lecture)

140

This lecture - Overview

• Evaluation research • Evaluation methods

141

Learning Outcomes

CLO1: Explain the fundamental principles of HF that can be used by civil and transport engineers to facilitate user- centred design

CLO2: Apply HF principles, methods and data to the design of road and traffic management systems

CLO3: Plan for the integration of HF into the design lifecycle of the road and traffic management system

Evaluation Research

143

Evaluation Studies

Human Factors specialists undertake research to: a) generate data about human capabilities, limitations and

characteristics that can be used to design things for human use b) evaluate things that have already been designed.

The purpose of evaluation is “to assess the effect of something”

Source: Sanders & McCormick (1987), p. 29

Evaluation Methods

145

Evaluation Methods

There are various Human Factors research methods that can be used to evaluate the effect of something that has been designed.

Criterion measures are the characteristics and behaviours measured that… “form the basis for judging the goodness of a design in an evaluation study”. They include:

• Performance measures e.g. number of data-entry errors; reaction time; time to complete set-up of equipment

• Physiological measures e.g. heart rate; respiration rate; visual acuity • Subjective measures e.g. self-reported opinions, ratings or judgments

about things like comfort, ease of use, mental workload, physical workload

Source: Sanders & McCormick (1987), p. 30

146

Evaluation Methods (2)

The criterion measures that you choose should be: • Reliable: stable when collected over time • Valid:

– Face validity – appear relevant to users – Content validity – the measure samples a measure relevant to the

domain of interest – Construct validity – the measure taps into the underlying construct of

interest. • Free from contamination: not be influenced by extraneous variables. • Sensitive: the measure should help to discriminate between good and

bad design

Source: Sanders & McCormick (1987), p. 32-32

147

Evaluation Methods (3)

There are several research tools and methods that are typically used in HF to evaluate a product. These include:

• Checklists • Focus groups • User Interviews • Observation techniques • Verbal protocol analysis • Task analysis • Experiments

Source: Popovic (1999). In Green & Jordan (1999); Sanders & McCormick (1987), p. 32-33

148

Evaluation Tools: Checklists

Used to identify: • deficiencies in designs • operations of a product/system • users’ needs for design modifications Do not provide rich data, however, about users’ experience

Source: Popovic (1999), p. 31. In Green & Jordan (1999)

Purpose Design Process Stage To define operations of a product/system and identify users’ needs

Early stages of design process and field testing

149

Evaluation Tools: Focus Groups

Help to assess the project use of a product or system

Users discuss issues related to usability of proposed designs

Helps identify issues important for users not taken into account by designers

Groups can be structured based on participants’ age, level of expertise, gender etc

Needs a facilitator and discussion guide (structured questions) Source: Popovic (1999), p. 31. In Green & Jordan (1999)

Purpose Design Process Stage To define operations of a product/system and identify users’ needs

Any stage of the design process

150

Evaluation Tools: User Interviews

Help identify users’ needs

Help to understand users’ culture

Help to understand contextual environment in which artifacts are going to be used

Provides designers with feedback on the acceptability of design concepts

Interviews can be unstructured, semi-structured or structured

Possible downsides – cost of interviewer; and interviewers might tell interviewer what they think might be “right” opinion.

Source: Popovic (1999), p. 32. In Green & Jordan (1999)

Purpose Design Process Stage To identify user’s needs Any stage of the design process

151

Evaluation Tools: Observation Techniques

Help to define the dynamics of the product/system/environment

Provide insights into difficulties users have while interacting with artifacts/products/systems

Observed directly (with manual recording of data) or by audio and video recording of user using the artefact/product.

Can be done in conjunction with verbal protocol analysis (i.e. users thinking aloud as they use the product/artefact etc)

Yield qualitative information that can be used to quantify user interactions

Source: Popovic (1999), p. 32. In Green & Jordan (1999)

Purpose Design Process Stage To define the dynamics of the product/artefact/system/environment

Final stage of the design process, during field test

152

Evaluation Tools: Observation Techniques (2)

Critical issue is to know how to observe and what to observe.

It’s important to specify: • The aim of the observational study • Scenario in which product/artefact will be used • Type of data to be collected • Presentation of data collected (how recorded) • Time available for observation • Recording tools (e.g. video; naked eye + Ipad)

Source: Popovic (1999), p. 32. In Green & Jordan (1999)

153

Evaluation Tools: Verbal Protocol Analysis

Also called the “think aloud” protocol

User talks aloud as they interact with product or artefact

Video recording is made of the user interacting with product or artefact

Experimenter generates transcripts which are segmented, interpreted and analysed

Helps designers get a better understanding of the principles and problems behind their design concepts

Source: Popovic (1999) In Green & Jordan (1999)

Purpose Design Process Stage To evaluate a design, users’ expertise levels and understand users’ concept of products

Any stage of the design process

154

Evaluation Tools: Task Analysis

Task analysis can be used to identify all user interactions with products and assess their usability

It involves identifying and documenting all user activities and tasks involved in interacting with an artefact/product

These activities and tasks are usually represented in diagrams and charts

The activities and tasks can then be used as the basis for defining specific activities and tasks that users are asked to perform during usability testing

Source: Popovic (1999) In Green & Jordan (1999)

Purpose Design Process Stage To define and evaluate operational procedures of the human/product/system

Concept development stage. Final design stage and field test

155

Evaluation Tools: Hierarchical Task Analysis

Here is an example of a Hierarchical Task Analysis (HTA)

Source: Stanton et al, (2013)

156

Evaluation Tools: Experimentation

Discussed on Day 3 of course

Used to assess whether some attribute of a product or artefact (e.g. road lane width) has an effect on human behaviour and performance

Can use an Independent Groups design, Matched Groups Design or Repeated Measures Design

Need to consider all the issues relating to experimental research discussed previously.

Source: Popovic (1999) In Green & Jordan (1999)

Purpose Design Process Stage To determine whether a design has an affect on human performance and behaviour for specific criterion measures

Final design stage and field test.

157

Evaluation Tools: Usability Testing

Is a structured evaluation process designed to assess the usability of a product or artefact against the following 6 criterion variables:

1. “Intuitive design: a nearly effortless understanding of the architecture and navigation of the site

2. Ease of learning: how fast a user who has never seen the user interface before can accomplish basic tasks

3. Efficiency of use: How fast an experienced user can accomplish tasks

Source: https://www.usability.gov/what-and-why/usability-evaluation.html

Purpose Design Process Stage To assess whether the artefact, product or system is easy to use, or “user friendly”.

Early stages of design process, when mock-ups are available

158

Evaluation Tools: Usability Testing

4. “Memorability: after visiting the site, if a user can remember enough to use it effectively in future visits

5. Error frequency and severity: how often users make errors while using the system, how serious the errors are, and how users recover from the errors

6. Subjective satisfaction: If the user likes using the system”

Source: https://www.usability.gov/what-and-why/usability-evaluation.html

Purpose Design Process Stage To assess whether the artefact, product or system is easy to use, or “user friendly”.

Early stages of design process, when mock-ups are available

159

Evaluation Tools: Usability Testing (2)

These are the basic steps in conducting a usability testing study:

1. Plan the study (users you want to test, questions you want to ask)

2. Recruit participants (5 is usually sufficient) 3. Design tasks – i.e. the ones that you want participants

to perform using the product or artefact 4. Run study (set up video, collect demographic data,

start tasks, take notes 5. Analyse the data for the 6 criterion variables

Source: https://www.hotjar.com/usability-testing/process-examples

160

Questions ?

Over to ….

161

Required Reading for Thursday Tutorial

Roberts, P. (2013). Impact of Roadside Advertising on Road Safety. Austroads Report AP-R420-13

ISBN: 978-1-921991-72-1

Focus on Section 9 of the report – the table, as you will be using the table during the tutorial with Mitch.

(Posted on Moodle)

162

Self-Directed Reading (Anytime)

PIARC (2015). The Role of Road Engineering in Combatting Driver Distraction and Fatigue Safety Risks.

(Will be posted on Moodle)

Prof. Michael Regan, PhD Research Centre for Integrated Transport Innovation

(rCITI) Room 112, Civil Engineering Building (H20)

E: [email protected]