Broadening Access to Statistical Data
by Improving User Interfaces
Ben Shneiderman
Founding Director, Human-Computer Interaction Laboratory
Professor, Department of Computer Science
Member, Institutes for Advanced Computer Studies &
Systems Research
University of Maryland, College Park
(ben@cs.umd.edu)
Human-Computer Interaction Laboratory
Interdisciplinary research community
- Computer Science & Psychology
- Information Studies & Education
(www.cs.umd.edu/hcil)
Scientific Approach (beyond user friendly)


Specify users and tasks
Predict and measure (theory-driven, hypothesis-testing)
time to learn
 speed of performance
 rate of human errors
 human retention over time


Assess subjective satisfaction
(Questionnaire for User Interface Satisfaction)


Accommodate individual differences
Consider social, organizational & cultural context
Design Issues

Input devices & strategies
Keyboards, pointing devices, voice
 Direct manipulation
 Menus, forms, commands


Output devices & formats
Screens, windows, color, sound
 Text, tables, graphics
 Instructions, messages, help



Collaboration & communities
Manuals, tutorials, training
www.awl.com/DTUI
usableweb.com
hcibib.org
useit.com
Library of Congress
Scholars,
Journalists, Citizens
Teachers,
Students
Visible Human Explorer (NLM)
Doctors
Surgeons
Researchers
Students
NASA Environmental Data
Scientists
Farmers
Land
planners
Students
LifeLines
Children as Our Design Partners
Zoomable User Interfaces (ZUIs)
Bureau of Census
Economists,
Policy
makers, Journalists
Teachers,
Students
Snap-Together Visualization

Allow
coordination
designers to
create novel
layouts that
combine
existing
visualizations

Allow users
to navigate
large datasets
conveniently
Census Statistics with Maps
Treemap - Stock market, clustered by industry
Universal Usability

Problem: Confusion, frustration, and remorse
dominate user experiences
Survey of 6000 users finds 5.1 hours/week wasted
 Incompatible files, interfaces, networks, hardware
 Poorly designed websites lacking accessibility policy


Solutions:
Raise user expectations
 Conduct research
 Provide practical tools & methods

Universal Usability

Problem: Confusion, frustration, and remorse
dominate user experiences
Survey of 6000 users finds 5.1 hours/week wasted
 Incompatible files, interfaces, networks, hardware
 Poorly designed websites lacking accessibility policy


Solutions:
Raise user expectations
 Conduct research
 Provide practical tools & methods

Skeptics corner
- Dumbing down
- Lowest common denominator
ACM Code of Ethics
In a fair society, all individuals would
have equal opportunity to participate
in, or benefit from, the use of
computer resources regardless of
race, sex, religion, age, disability,
national origin or other such similar
factors.
Internet Use by Education - 1998

Percent of U.S. Households Using the Internet
Total U.S., Rural, Urban, and Central City Areas
Falling Through the Net: Defining the Digital Divide www.ntia.doc.gov
Internet Use by Income - 1998

Percent of U.S. Households Using the Internet
Total U.S., Rural, Urban, and Central City Areas
Falling Through
the Net:
Defining the
Digital Divide
www.ntia.doc.gov
Technology variety: Support broad range of
hardware, software, and network access
1 to 100 range in processor speeds
286 486
Pentium
1 to 100 range in screen sizes
Palm
devices
30,000
Device Independence
Input: keyboard, speech,...
Output: visual, auditory,...
Conversion: Text-speech
Speech-text,...
Software Versions
Laptops
480,000
Large Desktop or Wall Display
3,840,000 pixels
1 to 100 range in network bandwidth
9.6K 56K
10,000Kbps
Compatibility
File conversion
Multiple platforms
User diversity: Accommodate different users
Language & Culture
Western, Eastern, developing...
Personality
Introvert vs extravert
Thinking vs feeling
Risk aversion
Locus of control
Planful vs playful
Skills
Computer newbie to hacker
Knowledge
Domain novice to expert
Disabilities
Visual, auditory, motoric, cognitive
Disabling conditions
Mobility, injury, noise, sunlight
Age
Young to old
Gender
Male or Female
Income
Impoverished to wealthy
Gaps in User Knowledge - Strategies
Bridge the gap between what users know and
what they need to know Online Learning
Design
Layered
Level-structured
Task-oriented
(evolutionary, phased)
Introductory tutorials
Getting started manuals,
Cue cards
Walkthroughs/Demos
Minimalist/Active
Training
Fade-able scaffolding
Training wheels
Minimalist
Online help
Context sensitive, tables of contents,
Indexes, Keyword search,
FAQs, Newsgroups, Chat rooms
Online communities
Customer service
Email
Phone
Help desks
ACM Conference on Universal Usability
Washington, DC
November 16-17, 2000
www.acm.org/sigchi/cuu
Broadening Access to Census Data
 Dynamaps
 Speeds
exploration of geographic data
 Facilitates comparisons & visibility of extremes
 Query
Previews
 Speeds
exploration of statistical data
 Facilitates study of distributions of data
 Table
browser
 Speeds
exploration of tabular data
 Facilitates understanding of data patterns
Dynamap
Dynamap
Dynamap
Broadening Access to Census Data
 Dynamaps
 Speeds
exploration of geographic data
 Facilitates comparisons & visibility of extremes
 Query
Previews
 Speeds
exploration of statistical data
 Facilitates study of distributions of data
 Table
browser
 Speeds
exploration of tabular data
 Facilitates understanding of data patterns
Problems with Browsing Large Online Tables
Universal Relation >> 1000 Tuples
e.g. 1997 US Economic
Census Data, for Counties:
more than 3000 Tuples
Hard to see overview,
gaps, and clusters or find
answers to questions
without many zero or
mega hit queries and large
unnecessary downloads
Conclusions






Visual presentation of data
Visual manipulation of data
Continuous immediate feedback on actions
Easily reversible actions
Error prevention instead of error correction
Better network performance:


No zero/mega hit queries
Loading only the metadata prior to query submission
Broadening Access to Census Data
 Dynamaps
 Speeds
exploration of geographic data
 Facilitates comparisons & visibility of extremes
 Query
Previews
 Speeds
exploration of statistical data
 Facilitates study of distributions of data
 Table
browser
 Speeds
exploration of tabular data
 Facilitates understanding of data patterns
Thomas Jefferson
I feel... an ardent desire to see
knowledge so disseminated
through the mass of mankind
that it may...reach even the
extremes of society: beggars
and kings.
-- Reply to American
Philosophical Society, 1808
ACM Conference on Universal Usability
Washington, DC
November 16-17, 2000
www.acm.org/sigchi/cuu
Human-Computer Interaction Laboratory
(www.cs.umd.edu/hcil)
References

Hert, C.A. (1999). Federal Statistical Website Users And Their Tasks: Investigations Of AvenuesTo
Facilitate Access: Final Report to the United States Bureau of Labor Statistics. Available at:
http://istweb.syr.edu/~hert/BLSphase3.PDF

Hert, C.A. (1998). FedStats Users and Their Tasks: Providing Support and Learning Tools: Final
Report to the United States Bureau of Labor Statistics. Available at:
http://istweb.syr.edu/~hert/BLSphase2.html


Haas, S.W. and Hert, C.A. (2000). Terminology Development and Organization in Multi-Community
Environments: The Case of Statistical Information. In Soergel, D. (ed.) Proc. 11th ASIS&T SIG/CR
Classification Research Workshop. Chicago, IL, November 12, 2000, p. 51-72

Hert, C.A. and Marchionini, G. (1997). Seeking Statistical Information in Federal Websites: Users,
Tasks, Strategies, and Design Recommendations: Final Report to the Bureau of Labor Statistics.
http://ils.unc.edu/~march/blsreport/mainbls.html


Marchionini, G. (1998) Advanced Interface Designs for the BLS Website: Final Report to the Bureau
of Labor Statistics. http://ils.unc.edu/~march/blsreport98/final_report.html

Marchionini, G. (1999). An Alternative Site Map Tool for The FedStats Statistical Website.
http://ils.unc.edu/~march/bls_final_report99.pdf

Marchionini, G. (2000). From Overviews to Previews to Answers: Integrated Interfaces for Federal
Statistics Report to the Bureau of Labor Statistics, June 30, 2000
http://ils.unc.edu/~march/bls_final_report_99-00.pdf