Human-Machine Interface Social Science Approach

Human Machine Interface
A Social Science Approach
The Eye, The Brain & The Auto 2013
September, 17, 2013
By
Lawrence Smythe EdD
Nissan Technical Center N.A.
Image Source:
http://www.henryford.com/body.cfm?id=58783

Request
This presentation is a general discussion regarding a broad range
of automotive HMI issues and does not necessarily reflect the
views and opinions of my employer, Nissan Technical Center, N.A.
This presentation also cites research from others and identifies the
sources of those findings. If any elements of this talk are used,
please be certain to cite the original sources if used in your writings
or presentations - Thanks

Autonomous Vehicles and HMI
Originally, Automobiles provided Transportation using
physical controls for operation, entertainment and comfort
Currently, Automobiles use similar physical controls but also
provide Electronic Device Interface Connectivity and Voice
Communications
In the Future, Autonomous Vehicles will Change the Needs
and Nature of Physical Controls and Electronic HMI because
the definition of Automobile and Driving is being redefined

Autonomous Vehicles and HMI
Key Point:
One of the many primary tasks regarding development of
autonomous vehicle HMI, is to create vehicles as a new type
of mass transit experience that allows autonomous driving by
accommodating Individual Differences in Expectations,
Learning Styles and Trust in Automation
Image Source: http://icsw.nhtsa.gov/safercar/ConnectedVehicles/

Some Current Activities Regarding Development of
Autonomous Vehicle Metrics and Technologies
SAE Safety & Human Factors subcommittee on Autonomous Vehicles
ISO TC22 SC13 WG8 HMI subcommittee on Autonomous Vehicles
Transportation Research Board’s June 2013 Workshop on Road
Vehicle Automation
My discussion will focus on the Top Four Issues raised
during the Human Factors Portion of last June’s TRB
Autonomous Vehicle Workshop at Stanford University

Top Four Questions from the Human Factors Portion
of the 6/2013 Transportation Research Board’s
Workshop on Road Vehicle Automation
1. How do we reengage the driver in Manual
Driving?
2. What does the user interface need to contain
to convey information?
3. What kind of misuse will occur and does
automation need to monitor the driver to
address this?
4. Should drivers be allowed to personalize
automation to accommodate their
tolerances?
Image Source:
http://www.vehicleautomation.org/
http://www.vehicleautomation.org/program/breakouts/human-factors-and-human-machine-interaction

Stanford TRB Top Four HF Questions: Discussion
1.How do we reengage the driver in Manual Driving?
What are the User Expectancies regarding Autonomous Vehicle hand-offs from
vehicle to driver and driver to vehicle? Are there special instances where
drivers characteristically ‘Take-Over’and would this process be systematically
different than ‘Emergency’Hand-off’s?
2.What does the user interface need to contain to convey information?
Will Autonomous Vehicle HMI systems be so different that we need to train
users? If not, what are the device and user stereotypes we can use to transfer
training from current to future HMI

Stanford TRB Top Four HF Questions: Discussion
3. What kind of misuse will occur and does automation need to monitor the driver
to address this?
What are the User Expectancies for ‘Normal Automated Driving’? What level
of ‘Human in the Loop’is required during different automated driving modes?
Can Misuse be designed out of the HMI operations?
4. Should drivers be allowed to personalize automation to accommodate their
tolerances?
What part or parts of the Automated Vehicle HMI would you personalize?
Colors and features are obvious, but personalization can also include choice of
different types of HMI based on personal preferences and capabilities.
Physical controls will have a defined range of accommodation, but
reconfigurable displays can be customizable so multiple interfaces are possible
for parts of the HMI

Background: Autonomous Vehicle Technologies will be Incremental in
that they will Build on Current Vehicle Systems, User Expectations and
User Experiences Using Technologies like……..
Trust, Epistemological Beliefs and Approaches towards Problem
Solving & Learning of Current Technologies Help Define User
Expectations of Future Technologies
All-around View Monitors
Backup-Collision Monitors
Blind Spot Warning & Lane Departure
GPS Navigation and Location Based Telematic Services
Adaptive Cruise Control & Adaptive Braking
Dynamic Vehicle Longitudinal and Lateral Control
Vehicle to Vehicle (V2V) Connected Car Technologies
Voice Control
Common Issues for this Discussion

What are Common Issues Regarding the Top Four Questions from
the TRB Human Factors Discussion on Autonomous Vehicles
Individual Differences between Users
A. Epistemologies: Unconscious Beliefs & Expectations
Individual Differences in Learning Styles
B. What are Some of things we know about Empowered
Learning verses Learned Helplessness?
Trust in Information and Automation
C. Some things we know about Trust in Automation

Epistemological Beliefs
Beliefs about Knowledge & Learning are Continuum and
Knowledge Domain Dependent
Example: Someone who has a Mature / Expert Understanding of
Baseball but has a Naïve / Novice Understanding of Tennis
Schommer, M., & Walker, K. (1995). Are epistemological beliefs similar across domains? Journal of Educational
Psychology, 87(3), 424-434
Factor Description Knowledge Belief – Learning Continuum
Simple
Knowledge
Knowledge
Integration
Isolated through Complex, Tentative, Interrelated Knowledge
Certain
Knowledge
Stability of
Knowledge
Unchanging through Constantly Changing Knowledge
Quick
Learning
Speed & Process
of Learning
Quick or Not-At-All through Occurs as a Continuous Process
Confidence
Meta-Cognition
Ability to learn
Fixed at Birth through Improves Over Time with Experience
Trust
Source of
Information
Reliance on Source / Authority through Develops Over time with Experience

Epistemological Beliefs Discussion
Smythe, L. : Development of mature beliefs coincides with development of
expertise through experience. Relationships therefore exist between novice
skills, naive beliefs, expert skills, and mature beliefs.
Individual differences in age, computer experience, educational
environments, learning styles, and personal epistemologies influence
learning outcomes.
Schommer-Aikins, M., & Hunter, R. (2002). Epistemological beliefs and thinking about everyday controversial issues.
Journal of Psychology, 136(1), 5-20.
Smythe, 2005, Measuring Learning Beliefs ProQuest Dissertation & Thesis, AAT3156904, ISBN 9780496894031
Schommer-Aikins, M. : People who believe in complex, tentative knowledge
were more likely to accept multiple viewpoints, they were also willing to
modify their opinions, withhold decisions until more information was
available, and acknowledge that everyday issues are complex and tentative

Smythe, L. : Since novice users have different information requirements
than experts, it is important to provide simplified knowledge frameworks
that also allow expert interactivity.
In this way, novices can use basic functions and develop expertise when
becoming confident (if) using a clearly understood knowledge framework.
If using a clearly understood knowledge framework, Expert users can
then creatively recombine functional elements of knowledge systems
because experts understand the interrelationships between functions.
Smythe, 2005, Measuring Learning Beliefs ProQuest Dissertation & Thesis, AAT3156904, ISBN 9780496894031

Temple, J. G. & Schmidt-Nielsen, A. (1998). Individual differences in word processing strategies. In R. R. Hoffman, M.
F. Sherrick, & J. S. Warm, (Eds.). Viewing psychology as a whole: The integrated science of William E. Dember. (pp.
205-226), Washington, DC: American Psychological Association.
Temple, J. G.: However, experience might not always lead to mastery because
some users improve their skills, and others are content with familiar, yet
inefficient methods.
It seems that inexperienced users who prefer simple choices find the complexity
of computers overwhelming and learn just enough to accomplish tasks.
However, inexperienced users who prefer complex choices are more likely to go
beyond minimum skills to explore increasingly complex computing.
Conclusion: Epistemologies are a range of Knowledge Domain
Dependent Beliefs so that everyone is a combination of naïve / novice to
mature / expert beliefs.
However, Learning style plays an important role in the development of
expertise which can lead to different levels of Empowered Learning

What are Some of things we know about Empowered
Dweck, C. S.: Some children favor what is termed an incremental theory of
intelligence: They believe that intelligence is a malleable, increasable,
controllable quality.
Others lean more toward an entity theory of intelligence: They believe that
intelligence is a fixed or uncontrollable trait.
A social-cognitive approach to motivation and personality.doi: 10.1037/0033-295X.95.2.256 By Dweck, Carol S.; Leggett, Ellen L.
Psychological Review, Vol 95(2), Apr 1988, 256-273
Are Learning Styles Static - Or do they Change over Time?
Percent of Subjects with each Theory of Intelligence Selecting each Achievement Goal
Goal Choice
Performance Goal Performance Goal Learning Goal
Avoid Challenge Seek Challenge Seek Challenge
Entity Theory (n-22)
50.0 31.8 18.2
Incremental Theory (n=41)
9.8 29.3 60.9

Grant, H.: Premed Students - Longitudinal Study: In a difficult premed course,
the impact of learning and performance goals depends on how they are
operationalized.
Active learning goals: (incremental theory) predicted active coping, sustained
motivation, and higher achievement in the face of challenge.
Performance goals: Ability-linked goals (entity theory) predicted withdrawal and
poorer performance in the face of challenge (but provided a Boost to performance
when students met with success)
(Intermediate Beliefs) Normative goals (wanting to perform better than others) did
not predict decrements in motivation or performance; and Outcome goals (wanting
a good grade) were in fact equally related to learning goals and ability goals.
Clarifying Achievement Goals and Their Impact. doi: 10.1037/0022-3514.85.3.541 By Grant, Heidi; Dweck, Carol S. Journal of
Personality and Social Psychology, Vol 85(3), Sep 2003, 541-553.
Conclusion: Results were Parallel to findings with younger students
who believed in Entity verses Incremental Theory

Plaks, J. E.: How do people respond to information that counters a
stereotype? Do they approach it or avoid it? Four experiments showed that
attention to stereotype-consistent vs. inconsistent information depends on people's
implicit theories about human traits.
Those holding an entity theory (the belief that traits are fixed) consistently
displayed greater attention to and recognition of consistent information, whereas
those holding an incremental (dynamic) theory) tended to display greater
attention to and recognition of inconsistent information.
This was true whether implicit theories were measured as chronic structures or
were experimentally manipulated. Thus, different a priori assumptions about
human traits and behavior lead to processing that supports versus limits
stereotype maintenance.
Person theories and attention allocation: Preferences for stereotypic versus counter-stereotypic information. doi: 10.1037/0022-
3514.80.6.876 By Plaks, Jason E.; Stroessner, Steven J.; Dweck, Carol S.; Sherman, Jeffrey W., Journal of Personality and Social
Psychology, Vol 80(6), Jun 2001, 876-893.

What are Some of things we know about Empowered Learning
verses Learned Helplessness?
This same study, measured which perceiver-type tends to focus on stereotype-
confirming information and which tends to focus on stereotype-disconfirming
information
In the presence of High Cognitive Load, Entity Theorists had Longer RT’s to
Stereotype Consistent Information whereas Incremental Theorists had Longer RT’s
to Stereotype Inconsistent Information
Consistent verses Inconsistent HMI Stereotypes Produce a Range of Responses

Epistemological Beliefs: Recap
Generally Speaking: Entity Theorists are people with Naïve / Novice
Beliefs whereas Incremental Theorists are people with Mature /
Experts Beliefs
Schommer, M., & Walker, K. (1995). Are epistemological beliefs similar across domains? Journal of Educational
Psychology, 87(3), 424-434
Factor Description Knowledge Belief – Learning Continuum
Simple
Knowledge
Knowledge
Integration
Isolated through Complex, Tentative, Interrelated Knowledge
Certain
Knowledge
Stability of
Knowledge
Unchanging through Constantly Changing Knowledge
Quick
Learning
Speed & Process of
Learning
Quick or Not-At-All through Occurs as a Continuous Process
Confidence
Meta-Cognition
Ability to learn
Fixed at Birth through Improves Over Time with Experience
Trust
Source of
Information
Reliance on Source / Authority through Develops Over time with Experience

What are Some of things we know about Trust in Automation?
Parasuraman, R.: The complexities of the operational environment combined with
the complexities of people may cause automation to be used in ways different
from how designers and managers intend.
Discovering the root causes of these differences is a necessary step to
provide automation that meets user needs and provides users with the authority
and decision-making tools required to use the automation to its best effect.
Humans and Automation: Use, Misuse, Disuse, Abuse Raja Parasuraman and Victor Riley Human Factors: The
Journal of the Human Factors and Ergonomics Society 1997 39: 230 DOI: 10.1518/001872097778543886
Effects of Warning Validity and Proximity on Responses to Warnings Joachim Meyer Human Factors: The Journal of
the Human Factors and Ergonomics Society 2001 43: 563 DOI: 10.1518/001872001775870395
Meyer, J.: Two Perceiver-types were found: Compliance refers to the
choice to take cautious actions in spite of system responses such as
‘Warnings” or ‘All Clear”. Reliance refers to users assuming that the
system is in a safe state or not when the indicators say so.

Lee, J. D.: People sometimes fail to use automation appropriately, because
they respond to technology socially.
Trust involves organizational, sociological, interpersonal, psychological, and
neurological issues regarding how context, automation characteristics, and
cognitive processes affect the appropriateness of trust.
Trust in Automation: Designing for Appropriate Reliance, John D. Lee and Katrina A. See Human Factors: The
Journal of the Human Factors and Ergonomics Society 2004 46: 50 DOI: 10.1518/hfes.46.1.50_30392
Merritt, S. M.: Individual differences affect perceptions of machine
characteristics, such that perceptions account for 52% of trust
variance
Not All Trust Is Created Equal: Dispositional and History-Based Trust in Human-Automation Interactions Stephanie M. Merritt and Daniel
R. Ilgen, Human Factors: The Journal of the Human Factors and Ergonomics Society 2008 50: 194 DOI: 10.1518/001872008X288574

(Continuing)
Merritt shows that there are two distinct types of trust
Dispositional Trust reflects trust in other persons (or machines) upon initially
encountering them, even if no interaction has yet taken place.
In contrast, History-based Trust is founded on interactions between the person
and another person or machine.
Conclusion: Although Dispositional Trust represents a relatively stable
construct, History-based Trust is a dynamic construct that adjusts as a function
of the person and machine’s cumulative interactions.
Not All Trust Is Created Equal: Dispositional and History-Based Trust in Human-Automation Interactions Stephanie M. Merritt and Daniel R.
Ilgen, Human Factors: The Journal of the Human Factors and Ergonomics Society 2008 50: 194 DOI: 10.1518/001872008X288574

Carsten, O. : Driving Simulator Study: Increased automation leads to increased
willingness to shift attention from driving to secondary tasks.
There are different effects of the two types of semi-automated driving, in that
drivers were more willing to engage in secondary tasks with lateral support than
with longitudinal support.
Control Task Substitution in Semiautomated Driving: Does It Matter What Aspects Are Automated? Oliver Carsten, Frank C.
H. Lai, Yvonne Barnard, A. Hamish Jamson and Natasha Merat, Human Factors: The Journal of the Human Factors and Ergonomics
Society 2012 54: 747 originally published online 17 September 2012 DOI: 10.1177/0018720812460246

Xiong, H. : Adaptive Cruise Control: Some drivers are internals
(conservative), and believe their behaviors are guided by their personal decisions
and efforts whereas externals (risky) believe their behaviors are guided by
external circumstances
Drivers in the conservative group stayed farther behind the lead vehicle than did
drivers in the moderate or risky groups. Risky drivers responded later to critical
events and had more ACC warnings issued.
The Conservative group trusted ACC more than the Moderate group and the Risky
group trusted ACC the least
Use Patterns Among Early Adopters of Adaptive Cruise Control, Huimin Xiong, Linda Ng Boyle, Jane Moeckli, Benjamin
R. Dow and Timothy L. Brown Human Factors: The Journal of the Human Factors and Ergonomics Society 2012 54: 722
originally published online 13 February 2012 DOI: 10.1177/0018720811434512

Domain Continuum
Epistemologies Naïve Intermediate Mature
Expertise Basic / Novice Intermediate
Advanced /
Expert
Common
Elements
Understands
primary concepts
and fundamental
methods; is able to
perform activity at
a foundational
level
Understands main concepts
and fundamental methods in
some detail; performs activity
above a basic level (e.g.,
could diagnose and solve
some problems, could show
someone with basic level;
skill how to improve
Understands
concepts and
methods in depth
and detail; is able
to perform
activity at an
expert level
Summary: Epistemologies
Beliefs about Knowledge and Learning are Continuum that Change
with Time, Experience and are related to Expertise, so…….
Autonomous Vehicle HMI must accommodate the Range of Naive to
Mature Knowledge Beliefs and these will change relative to Changes
in Expectations using Electronic Products
A psychometric examination of
Mission Essential Competency
(MEC) measures used in Air
Force distributed mission
operations training needs
analysis.
doi: 10.1037/h0094964
By Alliger, George M.; Beard,
Rebecca; Bennett Jr., Winston;
Symons, Steven; Colegrove,
Charles, Military Psychology,
Vol 25(3), May 2013, 218-233.

Summary: Learning Styles
People who believe intelligence is increasable (incremental theory ) pursue the learning
goal of increasing their competence, whereas those who believe intelligence is fixed
(entity theory) are more likely to pursue the performance goal of securing positive
judgments of that entity or preventing negative judgments of it.
In the presence of High Cognitive Load, Entity Theorists had Longer RT’s to Consistent
Information to their stereotypes whereas Incremental Theorists had Longer RT’s to
Inconsistent Information to their stereotypes
Autonomous Vehicle HMI must account for Information
Expectancies and that Learning Style Differences may Require
Giving Drivers a Choice of Preferred HMI
Generalized Learning Process
Familiar - Comfortable - Confident - Competent

Summary: Trust
People sometimes fail to use automation appropriately, because they respond
to technology socially.
Perceptions (about machines) account for 52% of trust variance
Increasing automation leads to increased willingness to shift attention from driving
to secondary tasks and the extent of attentional shift is a function of trust
People Respond to Automation Socially such that their Beliefs and
Perceptual Styles influence the way they Trust and Use Automation
Result: Autonomous Vehicle HMI must be viewed as a social experience
and that people will respond and communicate with it according to social
expectations

Conclusion
People approach using of automation socially and because of this, social learning theory like
Vygotsky’s Zone of Proximal Learning, Bandura’s Social Cognitive Theory, Brunswik’s
Probabilistic Functionalism and other similar approaches may lead to solutions for the Top
Four Questions from the Stanford TRB Workshop.
To accomplish the task of Designing Autonomous Vehicle HMI,
maybe the best approach is the oldest of approaches - Namely, a
Unified Theory of Psychology
Psychology: The empirical study of epistemology and
phenomenology. doi: 10.1037/a0032920 By Charles,
Eric P. Review of General Psychology, Vol 17(2), Jun
2013, 140-144.

Human-Machine Interface Social Science Approach

Recommended

Recommended

More Related Content

Similar to Human-Machine Interface Social Science Approach

Similar to Human-Machine Interface Social Science Approach (20)

Human-Machine Interface Social Science Approach