Analysis of Navigability of Web Applications for Improving Blind Usability

Presented by Lindsey Flash

April 22, 2008

Written by: Hironobu Takagi, Shin Saito, Kentarou Fukuda, and Chieko Asakawa

Published: ACM Transactions on Computer-Human Interaction September 2007

2

Why is This Important?

An ever increasing amount of functionality is becoming available on the web

An increasing legally blind population Currently 179,000 in Japan Currently 1.3 million in the U.S.

Advances in technology and competition have made many sites visually rich

Designing more usable web interfaces for everyone

3

Questions

Do blind users parse a webpage into subregions when they access a page?

How inaccessible are current online shopping sites?

How is the regulation for web accessibility contributing to increasing accessible online shopping sites?

How are blind users accessing the less accessible sites? What leads to failure?

4

Visual vs. Nonvisual Navigation

A short demonstration…

5

Simulating NonVisual Navigation

Animals Cat Dog Beagle Retriever Turtle Mouse Bird Parakeet Hawk Fruits Apple Red Delicious Granny Smith Orange Lemon Cherry Lime Grape States Ohio Texas Florida California Maine South Dakota Oklahoma Arizona New Hampshire Colors Blue Navy Blue Sky Blue Green Red Purple Indigo Lilac

6

Visual Navigation

Animals Cat Dog Beagle Retriever Turtle Mouse Bird Parakeet Hawk

Colors Blue Navy Blue Sky Blue Green Red Purple Indigo Lilac

Fruits Apple Red Delicious Granny Smith Orange Lemon Cherry Lime Grape

States Ohio Texas Florida California Maine South Dakota Oklahoma Arizona New Hampshire

7

Mental Models of Scanning

8

STUDY 1: Evaluation of Online Shopping Sites’ Navigability

Accessibility difficultiesVariety of small promotional “tools”

• Best seller rankings• Personalized recommendations• Comments from buyers

Strong domination of visually oriented business logic

• Screen real estate

Rapid changes of items

9

Nonvisual Navigability Analysis

Standard guideline: W3C’s Web Content Accessibility Guidelines (WCAG)

“Nonvisual Usability Visualization Method” Analyzes navigability based on graph

structures defined by heading tags and skip links

Calculates reaching time to each part of a page when using voice browsers

An important, and new, contribution of this paper

10

11

Measured Criteria

30 online shopping sites evaluatedUnited States 10Japan 10U.K 10

CriteriaReaching times to main contentRatios of pages with heading tagsRatios of pages with skip links

12

Results from Study 1

13

STUDY 2: Experiment on Voice-Access Behavior

Experiment 1Examines the entire process of

realistic online shoppingCompares completion times between

sighted and blind shoppers Experiment 2

Compares users’ behavior on the original pages and accessibility-enhanced pages

14

Experiment 1: Comparison of Shopping Processes between Sighted and Blind Users

Subjects asked to buy a DVD for a popular TV program

Average completion times:46.3 seconds for sighted users563.6 seconds for blind users

Blind participants more than 10 times slower than sighted

15

Experiment 2: Comparison of Original Pages and Accessibility-Enhanced Pages

Accessibility enhancements:Alternative texts added for imagesSkip links added at top of each pageItem numbers added to each itemHeading tags added to all pagesNo content removedOrdering unchanged

16

Experiment 2: Comparison of Original Pages and Accessibility-Enhanced Pages

No difference in results for reaching times to main contentParticipants lacked experience in

navigating accessible sites Participants spent significantly longer

time in the content areas Participants learned various

landmarks around the main content in each experimental session

17

Future Work

Consideration of a landmark-oriented model for nonvisual navigation

Simplification of navigation interfacesParticipants mainly used only four

commands for navigation in Study 2 Automatic suggestion of navigation

methods Integration of transcoding functions

18

Summary Conclusions

New regulations regarding accessibility have significantly contributed to better navigability in the U.K.

Blind users took more than 10 times longer to locate a specific search item compared with sighted users

Landmark-oriented nonvisual navigation model proposed to increase navigability

19

Questions?