Finding #1

There are tremendous differences among campuses  (combined data sets)

UW Campuses (2000)                                   

[percentages]                      
[average errors]                   

National Sample  (2000)

[percentages]
[average errors]

 

Finding #2

Differences among campuses are even more pronounced for specific data sets

 

    UW campuses (2000)

    Nationwide Sample

 

Finding #3

On the average, general campus pages have about the same degree of accessibility in all three data sets.

Table summarizing general campus data

   Bobby-Approved Pages
 UW 1999

48%  (SD=36)
 UW 2000

43%  (SD=36)
 National 2000

46%   (SD=32)

 

 

Finding #4

In 1999, library web sites within the University of Wisconsin System were not paragons of virtue; they tended to be less accessible than campus pages. 

Table summarizing UW data (1999)

   Bobby-Approved Pages
 Campus

48%  (SD=36)
 Library

31%  (SD=33)

 

[Campus and library data comparison--1999]

 

Finding #5

In 2000, library web sites within the UW system, on the average, were still not paragons of virtue. However, some overall improvement occurred.

Table summarizing UW data (2000)

   Bobby-Approved Pages
 Campus

43%  (SD=36)
 Library

40%  (SD=35)

 

[UW campus and library data comparison--2000]

 

Finding #6

On larger campuses (nationwide sample), library web pages tend to be more accessible than campus web pages.

Table summarizing National data (2000)

   Bobby-Approved Pages
 Campus

46%  (SD=32)
 Library

65%  (SD=32)

 

[Nationwide campus and library data comparison--2000]

 

Table summarizing Library Data

   Bobby-Approved Pages
 UW 1999

31%  (SD=33)
 UW 2000

40%   (SD=35)
 National 2000

65%   (SD=32)

 

 

Finding #7

While, on the average, UW library web pages showed some improvement, no improvement was found for UW campus web pages.

Table summarizing UW data (1999 and 2000)

 Bobby-Approved Pages
1999 2000
 Campus 48%   (SD=36) 43%   (SD=35)
 Library

31%   (SD=19)

 40%  (SD=35)

 

UW campuses data (1999 and 2000)

 

Finding #8

LIS Schools/Departments do not assume a leadership role in accessible web design.

Table summarizing National data (2000)

   Bobby-Approved Pages
 Campus

46%  (SD=32)
 Library

65%   (SD=32)
 LIS

19%  (SD=28)

 

[Percentages of Bobby-approved LIS data pages (nationwide samples, 2000)]                    

[Percentages of Bobby-approved campus, library and LIS pages (nationwide data, 2000)]

[Average errors per web page, by campus, library, and LIS pages (nationwide data, 2000)]

 

Finding #9

Most accessibility errors are due to the omission of alternative text for graphical elements (both simple images and client-side image-map hotspots).      

 

Finding #10

A high percentage of high-priority accessibility errors could be easily corrected.

      List of accessibility errors

      Easy to correct

  •  No alternative text for image
  •  No alternative text for hot spots on image map
  •  No title for frame

        Somewhat difficult to correct

  • No alternative text for applet
  • No alternative content for applet
  • No alternative contend for script containing important information

        Hard to correct

  • Page not readable and usable without frame-capable browsers
  • Server-side image without viable text-based alternative

 

[brief overview]  [Go to: accessibility page data]  [Go to: companion web site to my DEED 2000 presentation]