Voting User Interfaces Learning from Disabilities

1
Voting User InterfacesLearning from Disabilities

• Jonathan A. Goler
• Research Affiliate
• MIT Media Lab

2
Outline

• Motivation - are Full-faced systems worse than
  DRE for Reading Disabled Voters
• The Machines
• NY Study
• Results
• New UI Design

3
Motivation

• Law in NY requires full face machines
• Because of the large display, lots of things
  going on gt is it harder for reading disabled
  voters?

4
The Players
5
The Players
6
The Players
7
Study Design

• N96
• Located at West Side YMCA Polling location
• 4 Poll workers
• Reading evaluation in between voting

8
Reading Battery

• Nelson-Denny -reading
• WRAT - reading
• WAIS -IQ estimation
• Subjects were tested RD iff WAIS-ND or
  WAIS-WRAT gt 1

9
Populations
10
Results
11
Results

Time to Complete Ballot



0936

0824

0712


0600
R
D


R
D


0448
e
Co
n
tr
o
l
Co
n
tr
o
l


MC
0336

0224

0112

0000
M
a
c
h
in
e

LS
, V
200
0
, iVotronic

12
Results
13
Study Conclusions

• DRE improves completion rate by forcing voters
  through the process
• Normal voters do better on DRE
• RD voters do better on full faced
• Why? Orientation - FF is a grid

14
A New UI

• Full Faced
• But add the benefits of guiding

15
Prototype
16
Future Work

• Implement and Test new UI
• Provide data and suggestions in partnership with
  voting equipment developers

17
Acknowledgements

• Ted Selker
• Lorin Wilde
• Larry Thomas Lalita Krishnamurthy
• Megan Webb, Sarah Dionne, David Friend
• ESS and Brett Honig
• Funding Bennan Center for Justice
• Carnegie Corporation