Loading document…
Opening in Pages for Mac...
Your browser isn’t fully supported.
For the best Pages for iCloud experience, use a supported browser.
Learn More
Cancel
Continue
Human-Computer Trust:
The How and Why Behind Our Faith in Technology
ABSTRACT
Technology is advancing at increasing speeds each
day, from our own personal pocket-computers to self-
sustaining smart homes. While technology is and has
become an integral part of our lives, it is also shaping
how we interact with our envi
r
onment in a novel way.
In a way that us as humans have established a
relationship with the technology we use,
whether we
choose to or not, this phenomenon will continue to
surround us. T
his relationship between the human and
computer will be the focal point of this study.
To be
clear, this paper will not focus on how the human and
computer interact with each other but rather how and
what technology can do, or already does, to strengthen
a users confi dence and faith in a computer system.
Features of what constitutes trust in technology are
proposed as well as factors that influence and identify
why and what we trust in the human-computer
relationship.
Author Keywords
Interaction; Sociability; Trust Model; Human-robot
interaction; Human-computer interaction; Human-
computer trust
INTRODUCTION
Personal assistants such as Siri, Alexa, Google home
and even autonomous vehicles. They all rely on
human input to manage our personal life and control it
in a way that we depend on them for real life
assignments, similar to how some of us depend on
lists and calendars to organize our lives. Why do we
trust these lists and calendars though? Well, because
we
can
control them, it is not that we our trusting a
calendar to have the correct days, it is that we our
trusting our own input
into
these calendars
.
From this,
I raise question
s
of trust between the human and
computer;
In what way can a computer and/or
computer application be implemented in order to
instill a trustworthy connection with the user? How is
technology shaped to make users confi dent in it? What
aspects and features of technology contribute to
providing a sense of comfortability to the user?
How
can we ensure trust between technology and those
who use it?
Every day humans across the world use
technology in one way or another that has the
potential to benefi t them. Through establishing a
framework of trust that can be broken down into
components, we can maximize the potential use and
effectiveness of technology. Using fi ve factors:
Reputation, UI Quality, Understandability, Perceived
Usefulness and Communication Trust/Security I
analyze how each of these contribute to creating and
improving trust within technology.
Motivation
In the abstract, I stated that this paper will focus on
how and what technology can do to gain a users
confi dence. Confi dence can be said to be a measure of
trustworthiness and in this case, trustworthiness in the
computer system we use. This is incredibly important
in todays world, as E-commerce grows and
technology advances, maximizing our potential use of
technology requires that we have faith in it. Aside
from getting the most out of the technology we use,
human-computer trust enables factors of risk. Through
establishing trust in a computer system, we in turn
then have faith in the system to carry out any tasks we
may ask of it, which minimizes the perceived risk an
individual is taking using technology. This is
especially important when it comes to E-commerce.
Large companies such as IBM, Google and Apple all
rely on a system of trust to ensure communications
Joe Van Horn
University of Iowa
Iowa City
,
USA
v
anhornjoe8@gmail.com
•
License: The author(s) retain copyright, but ACM receives
an exclusive publication license.
This text fi eld is large enough to hold the appropriate release
statement assuming it is single-spaced in Times New Roman 8-
point font. Please do not change or modify the size of this text
box.
Each submission will be assigned a DOI string to be included
here
.
and actions taken by their users are secure, reliable
and confi dence instilling. [10]
On a wider scale, trust will and is becoming a major
factor for the future of technology. With robots such
as personal assistants that have the potential to
manage an individuals every need, questions of how
deep our trust in technology may go must be
answered.
This is important for two main reason
s
:
First, because the human life and a computer will
never work hand in hand until
trust
is established. For
example, would you trust Amazons Alexa to manage
and keep track of your professional career? Or even
more so, would you trust a self driving vehicle to get
you to work? Not unless we knew for
certain
that it
would never make a mistake. Unfortunately, that’s
currently not possible because nothing is ever a sure
thing. Innately, we trust our human selves to
do
that
sort of work for us. My question is, what does
technology need to do to in order to create a trusting
relationship with humans so that we can naturally
depend on it without worr
y? This is how the future
will progress and until there can be trust in putting our
lives in a computers hand, the ceiling is limited. That
constitutes the motivation behind this paper, a look
into what the future could be must start with what the
present is. By this I mean that in order to dig deeper
into human-computer trust at a high level, examining
the current features of our faith in technology is the
fi rst step towards a seamless connection between the
user and system.
Review of Literature
Research of this particular topic is broad, many papers
study solely the autonomous vehicle and how we will
begin to use them. In a study of assisted and
autonomous driving [2], the source explains the social
aspects of the road, how an autonomous vehicle will
interact with other drivers on the road. The primary
focus of this piece was on the interpretation of other
vehicles on the road from a human point of view and
then from a self-driving vehicle point of view. This
study was aimed at the future, with the assumption
being made that humans are already comfortable
using these vehicles, that the computer need not
establish a relationship with the human. In a study
focusing on personal assistant [1], a group had
decided to shine
light on the future of what personal
assistants(Alexa, Siri) could be and what will need to
be done to make them operate within the world. It
goes into detail regarding how a personal assistant, in
the future, will “act on our behalf in both an open
physical world and in cyberspace to negotiate with
other assistants, to make inquiries with people who
can perform actions or are thought to know the
answers to the questions at hand”. It then goes on to
say “such assistants would need to behave in [a]
fashion that people expect, per expectations of
cognition and with norms of action over time.” In
order for these personal assistants to “behave in
fashion”, they must fi rst know the person being dealt
with. The only way to do that is to establish a
relationship with them, to create a bond that makes the
user feel completely trustworthy of the actions they
may ask of the personal assistant. As for the present,
rather than what human-computer trust could be, a
study pertaining to a model of human-computer trust
in todays applications was published through the IEEE
Computer Society[9]. This research was focused
towards the interaction of a user. How does a user
interact with a computer that incline them to trust the
system? Using three main factors: interaction
intention, computer system trust and communication
trust, the researchers proposed a model to measure
how trust is affected through interaction. More so, this
study went into solutions for improving the trust a
user has in a computer system. Direct solutions such
as displaying trust measurements to the user,
providing recommendations which will hopefully help
the user and therefore increase their faith in the
system.
Research Questions
On the concept of human-computer relation; What
actions specifi cally cause a human to trust one
another? What can a computer do to effectively
establish a trust with the user?
There is no one correct answer for how trust should be
defi ned, part of it is confi dence, willingness to act on
behalf of the trustee and another factor, faith.
However, defi ning trust can be broken down into
many components, not all of which can be applied to
human computer trust though.
[9] One defi nition of trust in human-computer
interaction was defi ned as:
Computer System Trust:
“Computer system trust is defi ned as the degree of
trustworthiness of a computer system, which concerns
system quality; information quality provided by the
system; UI quality such as ease of use and its inartistic
trust solutions”. This defi nition led me to identifying
questions pertaining to what leads an individual to
conceiving a lasting faith in the technology they use.
Using technology requires that it work properly. A
useful application will be used to the extent that the
user desires depending on the applications capability.
[10] However, at what point will a user depend on a
technology, rather than simply use it for minor tasks?
Individuals will depend only on the technology they
trust, not those applications of which only provide
static content. Becoming dependent upon an
application requires it to be dynamic to the users
needs as well as intuitive. The goal is for the system to
make correct judgement based off users requests, but
in order to truly believe in those judgements, they
must be made as a result of the same psychological
procedure humans take when making judgements.[10]
(emotional reactions, intuitive responses)
During the time I spent researching and reading on
topics related to the one I am focusing on, many
interesting questions arose. In reading a study on the
future of personal assistants [1], the concept of the
devices being able to read emotion on a users face was
proposed. Which made me wonder how that could
improve the ability to trust the device. For example, If
Amazons Alexa could read the users face and actions,
thus becoming aware of the users emotional state, and
then show perceptual recognition of that, would it lead
to a stronger relationship? Would the user trust in what
Alexa says more? And in turn, trust the actions of
Alexa more? Sadly, this sort of emotional reading
hasn’t been perfected yet so there is no experimental
way to truly examine it. However, we can infer
certain aspects of trust from the way it’s implemented
in humans. What is it that causes one person to
completely trust another? We begin to trust someone
when they do something trustworthy. In a simple
form; If you tell a friend a secret and he or she honors
that for a period of time, you would likely then trust
that friend. So how could this process be implemented
into a computer and performed with the user? That is
the question at hand. [4] There was a recent [4] study
at New York University on the question of why do we
trust? It was stated that our trust in a stranger is
dependent on their resemblance to someone we
currently or previously have known. Could a smart
device somehow provide a user with a sense of
resemblance? A recent update to Amazons Alexa
implements a new feature which acts in a manner
similar to the ability to resembling their user. The
feature is called Alexa voice profi les. Using voice
profi les, Alexa is able to distinguish from one person
to another, recognizing the sound their voice and tone.
This is not exactly a feature of resemblance but it is a
start. Someone or in this case something that shows
resemblance to us means that they perform certain
actions that we fi nd similar to what we know, and
what we know is what we can trust. Alexa then would
be providing a user with a personalized experience
and over time, we would adapt to this experience and
Alexa in this way will be acting as if she knows who
she is speaking with, and if the user knows that too,
that is a feature of resemblance which can strengthen
trust.
Resemblance fi ts into the category of
understandability. Understandability is defi ned in this
study as the ability to analyze content, think about it
and intelligently deal with the objects presented.
Using understandability as one of the fi ve factors
contributing to trust, resemblance stands out. In
computing systems, applications and technology as a
whole. It can be assumed that users will almost always
feel most comfortable with the technology they can
resemble with most, those applications which they
have experience using, and with experience, comes
understandability.
Materials
To research trust and what it means between the user
and computer and more importantly how does it come
about. I conducted an online survey consisting of 19
questions that included demographics in the fi rst
quarter(anonymously) the rest were trust modeling
questions. Before releasing the survey to collect data,
I gathered information on how humans trust through
many research papers and journals, all of which were
published through a credible source such as the CHI
conference journals.
Participants
Following the creation of the survey, I decided to post
the survey online through the use of social
media(Facebook). This means that all of the
participants that took the survey had to be following
my social media accounts. Based off this it can be said
that I collected data randomly but within a select
group(only those who follow me on social media).
The majority (76
%
) of participants were students due
to the majority of my followers being students. Of
those students about 50
%
of them had claimed to be
business majors. In total, I ended up collecting
responses from 38 participants, 8 of which did not
complete the survey in entirety however. So I
excluded those results and my fi nal sample size came
to be 30 participants.
Participants ranged from ages 18 to 40+ years old.
93
%
of participants recorded that they were 18 to 25
years old. The remaining 7
%
logged as 40+ years of
age.
76
%
were
male
.
100
%
of participants have either completed a college
degree or currently in progress of completing a
college degree.
My hope was that the participant group would include
mainly those aging from 20-30 years old because they
will be the ones constructing and advancing
technology over the next couple decades.
Design
In designing the survey, I decided to use an online
survey software called
Qualtrics
which was provided
through the University of Iowa. The survey began as a
simple fi ve question survey in order to gain
demographic information of the participants, however,
after coming to the realization that it would be most
effi cient to make one large survey consisting of both
demographics and trust model questions. A survey of
19 questions, 5 being demographic and the other 14
pertaining to Human-Computer Trust was the fi nal
survey distributed to participants.
The survey was categorized into fi ve groups:
Reputation, Understandability, UI Quality,
Communication Trust/Security, and Perceived
Usefulness.
The defi nitions presented for these fi ve categories
were created using previous research [9][10]
defi nitions which were then combined/altered into a
list of fi ve factors used to estimate trust as well as to
seek the answer to which of the fi ve factors correlate
with trust the most.
Results
After gathering the data, I organized it into numerical
values. Each question and its following answer options was
assigned a number. For example if there were 5 answer
options for a question, they would be numbered 1-5 so that
I could create charts based off which answers participants
chose.
Reputation
Question 13 was presented to participants as an image
consisting of four different websites homepages. They were
asked to click on the interface they feel most comfortable
with. The four websites were:
FiveThirtyEight.com
,
TMZ.com
,
SoundCloud.com
, and
thegazette.com
. All of
these website have similar UIs but distinct enough to have
key differences. As you can see in the image above, the
bottom left website which is SoundCloud collected the most
clicks. Participants had felt most comfortable with
SoundClouds interface. This is due in part to the poor
design of the other interfaces such as poor consistency of
color on
TMZ.com
, a dramatic switch from black to white
is not
aesthetically
pleasing. More so, the main factors
contributing to this result is resemblance, site structure and
credibility. Not surprisingly, I expected SoundCloud to be
Reputation
Perspective of the public
opinion; brand, popularity.
Understandability
the ability to analyze
content, think about it and
intelligently deal with the
objects presented.
UI Quality
The feeling of use;
attractiveness; easy to
follow; entertaining
Communication Trust/
Security
The degree of
trustworthiness in
communications via a
computer system
Perceived Usefulness
The degree to which a
person believes that using a
particular computer system
would enhance his or her
use of the system
the most clicked on as it is the most popular of the four
websites. It also has a good use of hierarchical organization.
Question 19 asked the participants rank 5 categories based
on importance (Information Content and Quality, Branding
and Usability, Reputation, Understandability and Privacy/
security)
The below pie chart shows the percentage make up of how
important reputation was to the participants. While the
results don’t vary all that much, it can be seen that 50
%
either ranked reputation as somewhat important or not at all
important.
However, when the data was grouped by major,
it showed that Business and Computer Science majors were
the only majors to rank reputation as very important. This
could be due to their large use of computers, having a
popular computer may result in a higher confi dence of what
can be accomplished on the computer.
Perceived Usefulness
When asking iPhone users how dependent they are on IOS
(scale 0-100),
and whether
they would feel more confi dent
in the devices ability to handle tasks,
with features such as
fi nger or face recognition
implemented(scale:
a lot - not at
all)
,
the results
correlated with a high dependency
on
Apples
IOS
system.
This sort of result tells us that users of iPhones or Apple
products in general tend to depend largely on the those
devices and features such as the fi nger print scanner and
face recognition. These features represent a sense of
resemblance to the user, making them more confi dent in the
computer to handle tasks. This also connects to perceived
usefulness; having these features perceive the user into
thinking a device can do everything they need it to, without
actually knowing for certain.
Using the fi ve factors of trust proposed in the introduction
of this paper: Reputation, Understandability,
Communication Trust/Security, UI Quality and Perceived
Usefulness. Based on the results pertaining to each factor,
after counting the responses which are assigned to one of
the fi ve factors it was found that the sum of questions
answered that showed increased confi dence and/or high
importance resulted in the following rankings.
1)
Perceived Usefulness
1)
More than 50
%
of answers correlated with
with this factor of trust
2)
Content Information/Quality was the
greatest contributor
2)
Understandability
1)
Ranked last when asked to order question
#24 based off importance, however had the
highest results for question #22
3)
Communication Trust/Security
4)
UI Quality
5)
Reputation
In determining result correlations and fi nding extremities, it
can be seen that the results are very much based off the
responses of mostly college students. Students are a good
sample in this case because students are using computer
daily.
It is clear that the most important factor to establishing trust
between a computer and human is the usability of the
computer, or the perceived usefulness. Participants had
most often selected choices that designated with the use of a
computer system. Content information and quality ranked
fi rst when asking which aspect of a computer system is the
most important to them. Privacy and Security was the
second most important factor in improving confi dence and
or trust within a computer system.
Conclusion
Reputation(major)
Field
Business
Computer
Science
Health &
Medicine
Other
Very
Important
80.00%
20.00%
0.00%
0.00%
Important
50.00%
0.00%
50.00%
0.00%
Somewhat
Important
16.67%
33.33%
0.00%
33.33%
Not
Important
33.33%
0.00%
33.33%
33.33%
Not all
Important
66.67%
0.00%
0.00%
33.33%
27%
17%
23%
13%
20%
Very important
important
Somewhat important
Not important
Not at all important
While this may not be the most expansive research, a good
portion of quality information was pulled out and analyzed
through the use of a survey. Questions that made the
participants ask themselves why they trust their own
computer that they use everyday were proposed, through
connecting the answers to a set of fi ve categories that sum
to establish what trust means when it comes to technology I
was able to get insight into why we trust a computer, what
information matters that contributes to instilling confi dence
in the user so that computers and human may work
seamlessly.
Appendix
Related sources of research:
1.
Cohen, Phil, et al. “On the Future of Personal
Assistants .” On the Future of Personal Assistants , 2018.
Information Interfaces and Presentation.
2.
Brown, Barry, and Eric Laurier. “The Trouble with
Autopilots.” Proceedings of the 2017 CHI Conference on
Human Factors in Computing Systems - CHI '17, 2017,
doi:10.1145/3025453.3025462.
3.
Heiden, Remo M.a. Van Der, et al. “Priming Drivers
before Handover in Semi-Autonomous Cars.”
Proceedings of the 2017 CHI Conference on Human
Factors in Computing Systems - CHI '17, 2017, doi:
10.1145/3025453.3025507.
4.
“Why Do We Trust, or Not Trust, Strangers? The Answer
Is Pavlovian.” Sciencedaily.com, 29 Jan. 2018,
www.sciencedaily.com/releases/
2018/01/180129154004.htm+
5.
Levin, David. “Why Don't We Trust Self-Driving Cars?”
Research AI Research Seeks to Grow Trust between
Humans and Computers Comments,
www.bu.edu/
research/articles/ai-research-grows-trust-between-
humans-computers/
.
6.
Amazon Customer Website
www.amazon.com/gp/help/
customer/display.html?
nodeId=202199440&ref_=pe_1840220_274827100_MP
_AUCC_MX_HV_Sub_Fri_echo_crm_03.09.18_learn-
my-voice
.
7.
Luria, Michal, et al. “Comparing Social Robot, Screen
and Voice Interfaces for Smart-Home Control.”
Proceedings of the 2017 CHI Conference on Human
Factors in Computing Systems - CHI '17, 2017, doi:
10.1145/3025453.3025786.
8.
Chen, Xiuli & Starke, Sandra & Baber, Chris & Howes,
Andrew. (2017). A Cognitive Model of How People
Make Decisions Through Interaction with Visual
Displays. 10.1145/3025453.3025596.
9.
Yan, Zheng, et al. “A Research Model for Human-
Computer Trust Interaction.” 2011IEEE 10th
International Conference on Trust, Security and Privacy
in Computing and Communications, 2011, doi:10.1109/
trustcom.2011.37.
10.
Sousa S., Lamas D., Dias P. (2014) A Model for
Human-Computer Trust. In: Zaphiris P., Ioannou A. (eds)
Learning and Collaboration Technologies. Designing and
Developing Novel Learning Experiences. LCT 2014.
Lecture Notes in Computer Science, vol 8523. Springer,
Cham
11.
Grandison, Tyrone, and Morri Sloman. “A Survey of
Trust in Internet Applications.” IEEE Communications
Surveys & Tutorials, vol. 3, no. 4, 2000, pp. 2–16., doi:
10.1109/comst.2000.5340804
12.
Lyngbaek, Peter, and Dennis Mcleod. “
Trust in a
Specifi c Technology: An Investigation of Its components
and Measures
.” ACM Transactions on Information
Systems, vol. 2, no. 2, 1984, pp. 96–122., doi:
10.1145/521.357415.
Survey:
Q1 How old are you?
18-20 (1)
21-25 (2)
26-30 (3)
31-40 (4)
40+ (5)
Q2 Ethnicity: Please specify your ethnicity.
White (1)
Black or African American (2)
Native American or American Indian (3)
Asian/Pacifi c Islander (4)
Other (5)
Q3 What is the highest level of school you have completed
or the highest degree you have received?
Less than high school degree (1)
High school graduate (high school diploma or
equivalent including GED) (2)
Some college but no degree (3)
Associate degree in college (2-year) (4)
Bachelor's degree in college (4-year) (5)
Master's degree (6)
Doctoral degree (7)
Professional degree (JD, MD) (8)
Q4 What is your sex?
Male (1)
Female (2)
Q5 Please indicate your occupation(s):
Student (1)
Management, professional, and related (2)
Service (3)
Sales and offi ce (4)
Farming, fi shing, and forestry (5)
Construction, extraction, and maintenance (6)
Production, transportation, and material moving (7)
Government (8)
Retired (9)
Unemployed (10)
Other (11)
Q
6
What is your major?
Business, Management, Marketing (1)
Parks, Recreation, Leisure and Fitness Studies (2)
Social Sciences (3)
Communication, Journalism (4)
Computer Science, Math and related Sciences (5)
Health and Medicine (6)
Other (7)
________________________________________________
Q7 Do you own a smartphone?
Apple iPhone (1)
Samsung (2)
LG (3)
Motorola (4)
Google (5)
Other (6)
I do not own a smartphone (7)
Q8 Do you depend on apples IOS(operating system for
your iPhone, mac etc..) to the same extent you depend on
yourself to complete tasks?
Q
9
Do you allow your devices to save payment information?
Yes (1)
Occasionally (2)
No (3)
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
1
0
0
Please move the slider
accordingly: 0 = Not at all,
100 = yes, I would be lost
without it. (1)
Q1
0
Given the following task:
Schedule a reservation for a
popular restaurant downtown
.
Which method would you be most confi dent with? Please
rank them accordingly.
Q1
1
Technology today manages to obtain large amounts of
data about individuals, such using your fi ngerprint scan or
even your face to access/unlock your phone. Would having
a feature like these increase your confi dence in the device's
capability to handle and complete tasks?
A lot (1)
A moderate amount (2)
A little (3)
None at all (4)
Q1
2
Have you seen HTTPS (as pictured below) in your web
browser, meaning a secure connection? If so, how does this
make you feel about your internet security?
0 (0)
1 (1)
2 (2)
3 (3)
4 (4)
5 (5)
6 (6)
7 (7)
8 (8)
9 (9)
10 (10)
Q1
3
Please click on the interface you feel most comfortable
with:
Q1
4
Check all that apply
Using computers enables me to accomplish tasks more
quickly (1)
Using computer makes my job easier (2)
Computers increase my job productivity (3)
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
1
0
0
Calling the restaurant via
telephone (1)
Walking into the restaurant
beforehand (2)
Scheduling it on their
website (3)
Asking a personal assistant
to schedule it (Alexa,
Google Home, etc…) (4)
Q1
5
Please choose from the following:
I like computer systems developed by famous vendors
(1)
Popular brands are a must have (2)
If the computer works well, the brand doesn't matter
(3)
Q
16
Please select the option(s) you relate with most
I use computer systems because they are designed
with sound security (1)
I use computer systems because it can preserve my
privacy (2)
I use computer systems because it is reliable and free
of problem (3)
I use computer systems because it provides free of risk
solutions (4)
I use compute systems because I feel safe using them
(5)
Q
17
Choose all that apply:
Using computer systems enables me to identify with
other communication entities (1)
The situation and purpose of a communication would
influence my willingness to use a computer system (2)
Using a computer system helps me know other people
in a good way (3)
Q
18
Please choose one from the following:
I use computer systems because the user interface is
attractive (1)
I use computer systems because it's user interface is
easy for me to follow (2)
I use computer systems because interacting with them
is fun (3)
Q
19
In choosing which applications/websites to trust or
distrust, rank the following in importance to you
______ Information Content/Quality (1)
______ Branding and Usability (2)
______ Reputation (3)
______ Understandability (4)
______ Privacy/security (5
