Usability and the Evolution of Technology

1
Usability and the Evolution of Technology
or, You Shouldnt Have to Read a User Manual to
Ride an Elevator!

• Thomas S. Tullis, Ph.D.Senior Vice President,
  Human Interface DesignFidelity
  Investmentstom.tullis_at_fidelity.com

November 3, 2005Philadelphia, PA, USA
2
A True Story

• Recently I was running late for a meeting at a
  Fidelity office in another building in downtown
  Boston.

A tall building that I go to pretty rarely.
Image from Wikipedia.org
3
The Elevators
4
I Get in the Elevator
5
Where are the Floor Buttons?!
6
Finally Back at the Lobby
7
The Elevator Saga
8
The Elevator Saga
9
The Elevator Saga
10
The Elevator Saga
11
The Elevator Saga
12
The ElevatorSaga
13
The Elevator User Guide!
14
The Elevator User Guide!
15
The Elevator User Guide!
16
The Tear-off Reference Card!
17
The Elevator Saga
18
The Elevator Saga
19
The Elevator Saga
20
The Elevator Saga

• How did we get to this point?
• Lets look at the evolution of the elevator

21
Otis 1861 Elevator Patent
22
First Commercial Passenger Elevator
1857 Haughwout Building in New York City, a store
that sold china, glassware, etc. The elevator was
powered by a steam engine!
23
Elevators Enabled Skyscrapers
1932 View of New York City (from Library of
Congress)
24
How Did You Use Early Elevators?
25
Elevator Operators, WWI
26
Elevator Operators, 1920s
27
Elevator Operators, 1932
28
Elevator Operators on Strike!
1930s Elevator Operators Strike, New York City
29
People Stranded!
1930s Elevator Operators Strike, New York City
30
My Memories of Manual Elevators in the 1950s
31
Elevator Operators
32
Autotronic Elevator
First operator-less automatic elevator was
installed in 1950 in the Atlantic Refining
Company Building in Dallas, TX.
33
Autotronic Elevator Ads, 1955
How would my tenants react to operatorless
elevators? This question is uppermost in the
minds of building managers who are concerned with
todays spiraling operational costs. Why not ask
your tenants? Otis Autotronic elevators give
tenants a sprightly feeling of independence.
Riders simply step into the car and press buttons
for the floors they want. Everything else is
completely automatic. Weve found that tenants
like the idea of self-service elevators. They
push buttons for each other. They tell new riders
what to do. Everybodys friendlier.
34
Transition Period Operators Push Buttons for You!
35
Very Few Manual Elevators Now
Dale Collier retired on Feb. 4, 2005, from his
12-year job as an elevator operator in Appleton,
Wisconsin.
36
Evolution of Elevators

• When a new technology is first introduced, its
  often only the select few who directly interact
  with it.
• Elevator operators
• Evolution to more automation and more
  self-service.
• Is it a better user experience?
• What about for people with disabilities?
• Technological mutations like the
  destination-based elevators
• Maybe what we will see everywhere in 10 years?

37
Evolution of Information Technology
A scribe working in a Medieval scriptorium.
38
Illuminated Manuscripts

• Strikingly beautiful
• Created by a small number of highly skilled
  artisans
• Extremely time consuming (years to create)
• Accessible by very few

39
The Printing Press
40
The Gutenberg Bible
41
Books

• Enhancements in printing presses eventually led
  to the type of books we know today.

• More utilitarian.
• Created by more people but still relatively few.
• Far less time-consuming to create.
• Accessible by more, often through libraries.

42
Dime Novels

• An interesting development in the late 1800s

• Not exactly fine literature, but very popular.
• Opened up reading to an entire generation of
  youth.
• Low cost (5 or 10 cents) made them accessible to
  a very broad range of users.
• Predecessor of todays paperbacks.

43
Electronic Books

• Can access content in many different ways.
• Can be inexpensive or free.
• Can be very portable.
• Accessible by an extremely broad range of users,
  including those with visual impairment.

44
Community Authoring Blogs

• Now we have not only widespread access to
  information, but also much broader participation
  in the creation of it.

45
Some Themes

• As a technology evolves we often see
• A move away from only the high priests having
  direct access
• Greater access by more people
• More diverse users
• More mobility or ubiquity
• Shifts in usability (up, down, sideways!)
• We have an opportunity to influence that!
• Lets look at one specific issue
• User authentication

46
User Authentication

• With the earliest technologies, authenticating
  users was not really an issue
• Elevator operators also acted as gatekeepers
• If you could get physical access to books, etc,
  you were generally allowed to use them.
• Now we have
• Card readers, keys, and codes for accessing
  special floors via elevators.
• A dizzying array of passwords to access
  computers, networks, websites, etc.

47
User Authentication

• Even with the earliest computers, user
  authentication was not an issue.
• If you (or your deck of punch cards!) could get
  physical access (controlled by the high
  priests) and identify your account, that was it!
• Even with the early personal computers, it was
  simply not an issue.
• It was when access to systems remotely became
  more commonplace that authenticating the user
  became important.

48
Passwords

• If you create strong, secure, passwords like
  youre supposed to
• Characters, digits, and special characters
• Mixed case
• Changed regularly
• Then youre probably going to
• Forget them, or
• Write them down and stick them under your
  keyboard!
• Can we look at user authentication from a very
  different perspective?

49
Authentication via Personal Photos

• Lets take a user perspective and capitalize on
  something that most people are very good at
• Recognizing pictures
• Especially pictures that are meaningful to them
• Maybe we can even turn authenticating into
  something thats actually fun!

50
Authentication via Personal Photos
When registering with the system, users provide
some photos that are of personal significance to
them, but which others cannot readily associate
with them. At login time, the user has to pick
their photos from among highly similar
distractors.
Demo!
51
Very Preliminary Results

• Very early pilot testing with users has shown
  that
• Users have absolutely no trouble recognizing
  their own photos among even highly similar
  distractors.
• Accuracy rate has been over 99 correct.
• Even others who know them quite well cant come
  close to breaking in as them.
• Users are surprisingly fast at recognizing their
  own photos.
• They report that its actually enjoyable!

52
Authentication via Personal Photos

• Still lots more design and testing to do
• Guidelines about photos, mechanism for providing
  them, assigning keywords, using the keywords to
  match up with likely distractors, etc.
• Exact numbers of personal photos required, number
  of distractors, number to recognize for
  authentication, etc.
• An example of something thats probably 2-3 years
  away from possible implementation.
• But a way in which we might be able to influence
  the evolution of user authentication.

53
The Most Recent Evolutionary Step

• Now lets turn to the most recent huge step in
  the evolution of information technology

54
The Web

• In an amazingly short time, the web has
  revolutionized how we access information and how
  we conduct business.
• For example, today the web accounts for over 90
  of our transactions with our customers at
  Fidelity.
• The web has the potential for opening up quick
  and easy access to information and transactions
  for an incredibly large and diverse group of
  people.
• Including many who have been marginalized by
  earlier technologies.

55
The Myth of the Average User

• Lets take a look at a representative sample of
  100 people.
• For simplicity, Im going to just look at the
  United States.
• Data primarily from
• 2000 U.S. Census
• 2005 Pew Internet Research report
• Going to look at characteristics that might
  matter in terms of interacting with the web
• Gender, age, education level, disability,
  language spoken at home, type of access to the
  Internet

56
100 Representative People
57
100 Representative People
58
Gender

• Does gender really matter when it comes to how we
  interact with the web?
• Maybe

59
Our Eye-Tracking Studies
60
Eye-Tracking Studies Sample Data
61
Eye-Tracking Studies Sample Data
62
Eye-Tracking Studies Sample Data
63
FIDO Study

• Investigated what kind of features and
  information users wanted on the homepage of their
  ideal financial services site.
• Gave them building blocks based on elements
  from our Fidelity.com homepage and the homepages
  of five of our competitors.

Presented at UPA 2004 Conference Tedesco,
Chadwick-Dias, and Tullis.
64
FIDO Study
65
FIDO Study
66
FIDO Study Results

• Men tended to put more news items on their
  homepage

67
FIDO Study Results

• Men also tended to put more than one quote
  field on their homepage!

68
FIDO Study Results

• Women were more likely to add elements related to
  special offers or promotions
• Interestingly, about 50 of the women put this
  specific element somewhere on their page

63 of women, 21 of men
So gender really might make a difference, in
terms of what users want from a website and how
they process it.
69
100 Representative People
Does age make a difference in interacting with
the web?
70
Web Usability and Age

• Conducted two studies using a prototype benefits
  site.
• Users over a wide range of ages (20s to 80s).
• Controlled for web experience (length of web use,
  frequency of use, variety of sites)
• Measured performance on the same set of tasks.

Presented at CUU 2003 Conference
Chadwick-Dias, McNulty, and Tullis
71
Web Usability and Age

• Modified the prototype for the second study to
  address issues we saw older users having in the
  first study, e.g.
• Cautious clicking
• Problems detecting where to click
• Understanding complex pages
• Problems with web or technical terminology (e.g.,
  windows, home, login)

72
Web Usability and Age
Age appears to have an impact on a users
performance on a website, but changes targeted
for older users helped everyone.
73
Web Usability and Age

• More recent study looked at a possible technique
  to address cautious clicking

74
Web Usability and Age
Having the help made no difference for people
under 65 but significantly helped people 65 and
over.
75
100 Representative Users
Does education level make a difference in ones
interaction with the web? Probably in indirect
ways, through overall literacy and subject-matter
literacy.
76
Financial Literacy

• In the study described earlier, with the pop-up
  help boxes, we also measured their level of
  financial literacy using a 20-question quiz,
  e.g., do they know the difference between a stock
  and a bond?
• Looked at the results of the study for those who
  scored
• lt 70 correct
• gt 70 correct

77
Financial Literacy
The help made no difference for those who scored
higher. It significantly helped those who scored
lower.
78
100 Representative People
About 1 out of every 5 Americans has some type of
disability that interferes with their everyday
living or their work.
79
The Drexel Homepage
80
How It Looks if You Have Cataracts
81
Wikipedia Article with Cataracts
82
But With Enlarged Text
83
Our Study with Low-Vision Users
84
Our Study with Blind JAWS Users

• Whats it like to listen to a website with a
  screen reader? (video clip)
• First conducted focus groups with blind users to
  learn more about how they use the web and what
  would make things better for them.
• Implemented several features in a prototype
• Page overviews
• Access keys
• More use of headings

Video clip
85
Our Study with Blind JAWS Users
A commonly used feature of JAWS is the Links List
dialog box.
86
Our Study with Blind JAWS Users

• Still analyzing the results, but some preliminary
  observations
• Page overviews are helpful but need to be short.
• Access keys are helpful but should only have a
  few.
• Having lots of links on a page is problematic.
• The importance of the first letter of the link
  text!
• The importance of good use of heading levels.
• The importance of a good site search feature.

87
100 Representative People
Anyone who has made the transition from dial-up
to broadband knows that Internet access speed
makes a real difference in your use of the web.
88
Recent Study of Connection Types

• Conducted by Pew Internet Research (Susannah Fox,
  2005)

89
Percent Online by Education Level
90
Percent Online by Age
91
100 People
Obviously someones preferred language can have
an impact on how they use the web and what they
need from websites (e.g., versions in their
language).
92
100 Representative People
So who is your average user?
93
My Stunning Conclusion from 6 Years of Graduate
Training in Human Factors and 30 Years of Work
Experience in the Usability Field

• People are different!

94
What Does This Mean?

• Does this mean we have to build different
  versions of our websites for all these various
  demographic groups?
• Not at all!
• It means we need to build in intelligent and
  easy-to-use features that allow the user to
  customize the interface to their needs and
  desires.
• And once they do this for one website most of the
  features should carry over to others!

95
A Concluding Analogy

• Im currently building a pool shed by our new
  pool something Ive never done before.

96
A Concluding Analogy

• Doors are a standard height 80
• Thats not the 95th percentile of the population
  height its more like the 99.9th percentile.
• Architecture has evolved to a level of maturity
  where they have
• Standards
• Standard building blocks
• And its much cheaper to make the doorway 80
  high from the beginning than to go back later and
  retrofit one in!

97
A Concluding Analogy

• We need to make the doors to our websites taller!
• Technology will continue to evolve, whether we
  like it or not.
• Usability people need to take an active role in
  making the right kind of evolutionary changes
  happen.