accessibility of spanish university web =============== accessibility of spanish university web ===================

Accessibility
of Spanish university
web

===============
ACCESSIBILITY OF SPANISH UNIVERSITY WEB
=======================================
In little more than a year from now, recent legislation will require
all Public Administration websites—as well as all other publicly
funded websites—to meet the criteria for accessibility. Included among
them are University websites, the contents of which fall considerably
short of being fully accessible to all students.
The purpose of this report is to shows just how compliant a
representative sample of Spanish university websites is in terms of
meeting the basic conditions of accessibility. To that end, a novel
approach was designed to include both the technical accessibility
analysis based on the Web Content Accessibility Guidelines 1.0 drawn
up by W3C/WAI as well as a usability and accessibility assessment
based on feedback from the users themselves. Thus, the technical
analysis, carried out by a team of experts in designing fully
accessible web pages, has been complemented by an assessment in which
users with varying types and degrees of disability provided feedback
on the difficulties they found while navigating through the university
web portals.
Introduction.
-------------
This report for the Info-accessibility Observatory, set up by the
Discapnet portal to promote web accessibility for everyone, is the
first in a series of studies on the accessibility of different types
of web portals. The choice of Spanish university websites as the type
for this initial report reflects how important it is for students,
faculty, and researchers to be able to access the net without
obstacles, whether in search of information or the growing number of
services being provided on the net. The study was undertaken by the
Department of Accessibility at Fundosa Teleservicios.
Before launching into the results, however, a word should be said
about what is meant here by “web portal accessibility.” It can be
defined as a set of technologies, application rules and design which
facilitates the use of websites in accordance with the guidelines of
“design for everyone.” An understanding of the basis underlying the
concept of web accessibility can be seen in the following cases,
offered for illustration purposes:
*
Totally blind users may be using a screen reader to access browser
content, either by hearing it read aloud on speakers or reading it
with their fingertips on a Braille display.
*
Users with poor eyesight, partial eyesight, or color blindness who
use systems for magnifying the screen may need to enlarge the font
size of the text, or may require higher color contrast between the
foreground and the background.
*
Users with limited motor skill ability in their hands may be
unable to use a mouse, and thus access the net exclusively through
the keyboard or a voice recognition program allowing them to
navigate by giving voice commands.
*
Deaf users may require text alternatives to multimedia sounds
elements on the web page.
*
There are also users whose Internet connections are very slow, or
who connect through small-screen handheld devices such as PDA’s
and cellular phones who would benefit from accessibly designed
websites.
On the net, just as in the brick-and-mortar world, any design that is
heedless of accessibility issues creates needless barriers and
hardships. In contrast, as Jakob Nielsen points out in his book
Usability in Website Design, an accessible design is closely related
to the overall usability of the site, and enhances the effectiveness,
efficiency, and satisfaction of their experience for all users
regardless of any functional limitation they may have.
Legislation in Spain has tried to promote criteria that foster
universal accessibility. The Constitution, drafted in 1978, designates
one of the State’s responsibilities as that of “promoting the
conditions so that freedom and equality for the individuals and the
groups to which they belong are real and effective; removing the
obstacles that impede their fulfillment, and facilitating the
participation of all citizens in public, economic, cultural, and
social life” (art. 9.1). Much more recent legislation includes efforts
to guarantee full accessibility to the web. Such is the case with Law
34/2002 of July 11 on Services of the Information Society and
E-business, the Fifth Additional Disposition of which sets the
deadline at December 31, 2005 for all Public Administration and
otherwise publicly funded websites to be accessible. Similarly, Law
51/2003 of December 2 on Equal Opportunity, Non-discrimination, and
Universal Accessibility for the Handicapped states in its Final
Seventh Disposition the basic conditions for accessibility and
non-discrimination in accessing and using the technologies, products,
and services related to the Information Society and social
communication media, with deadlines ranging from two to ten years,
depending on the type and origin of the information concerned.
The European Union has also initiated measures regarding website
accessibility, including the European Parliament Resolution on
Communication from the Commission “e-Europe: accessibility of public
websites and their content” (April 2002), which advocates developing
an internet accessible to all citizens. Point 32 of the Report to the
European Parliament on the Communication of the Commission of
Industry, Foreign Trade, Research, and Energy of April 24, 2002
underlines that “for websites to be accessible, it is essential that
they meet the double-A level, and that they fully comply with all the
verification points of priority 1 and 2 from the WAI Guidelines.”
Other studies have been done on the accessibility of university
websites. C. Egea García’s “Server Accessibility in the Public
Administration” (1998) used a sample including three Spanish
university websites (the Complutense of Madrid, and the General
Studies of Valencia and Murcia) and showed that none of the three met
even minimum requirements of accessibility. Later, a 2002 study by M.
Térmens, Ribera, M. and Sulé, A. titled “Level of Accessibility on
Spanish University Websites” found that there was still much ground to
be covered, given that only 16 of the 256 pages analyzed complied with
the minimum requirements.
Note: the field work on which the present report is based was carried
out between August 5-30, 2004 (technical verification) and the second
week of September (user feedback tests). Some portals may since have
modified their sites, thus affecting their subsequent level of
accessibility.
Sample Selection.
-----------------
The basic criteria for selecting the sample of Spanish university web
portals used in this study are based on the need to reflect both its
relevance for the student as well as the diversity of the educational
sector: differences in institutional size, location, public or
private, on-campus vs. distance learning, etc. With these factors in
mind, the selection was made of:
*
The seven largest universities in terms of enrollment (more than
50,000, all public, one giving distance learning).
*
Four mid-sized universities (enrollment between 15,000 and 50,000
students; three public and one private distance learning college).
*
Two small universities (fewer than 15,000 enrolled; both private).
The criterion for variety in geographical location is met, since the
sample is spread out among 8 Autonomous Regions, both single-province
regions as well as regions consisting of various provinces, of
differing size and population density.
In terms of student enrollment, the university portals selected
represent 41.6% of all Spanish university students enrolled in the
2002-2003 academic year (619,555 students).
In addition to the portals of the universities themselves, it was
considered valuable to include two websites hosting general-interest
university information. Thus, both the Council of Rectors of Spanish
Universities (CRUE) portal as well as the Ministry of Education’s
University Information (MECD) website were included in the sample.
Table 1 shows the selection of portals used in this study, with
relevant data for their inclusion in the sample.
Table 1- Portals making up the sample in this study
UNIVERSITY PORTALS
University
Abbr.
Sector
Enrollment
Region
Universidad Nacional de Educación a Distancia1
UNED
Public
128,729
National
Universidad Complutense de Madrid
UCM
Public
88,216
Madrid
Universidad de Sevilla
USE
Public
67,365
Andalusia
Universidad de Granada
UG
Public
58,009
Andalusia
Universidad de Barcelona
UB
Public
57,219
Catalonia
Universidad del País Vasco
UPV
Public
51,665
Basque Country
Universidad de Valencia (Estudios Generales)
UV
Public
50,896
C. Valenciana
Universidad de Valladolid
UVA
Public
31,232
Castilla y León
Universitat Oberta de Catalunya1
UOC
Private
23,868
Catalonia2
Universidad de Las Palmas de Gran Canaria
ULPGC
Public
20,837
Canary Islands
Universidad de Alcalá de Henares
UAH
Public
18,808
Madrid
Universidad de Deusto
UD
Private
12,050
Basque Country
Universidad de Navarra
UN
Private
10,661
C. Navarra
GENERAL REFERENCE PORTALS
Name
Abbreviation
Council of Rectors of Spanish Universities
CRUE
University Information, Ministry of Education, Culture, and Sports
MECD
1 Universities offering distance learning degrees.
2 The Universidad Oberta de Catalunya offers study programs outside
the Catalonia Region but, unlike UNED, it does not host any support
centers outside Catalonia.
Checking for accessibility: technical aspects and user feedback.
----------------------------------------------------------------
Consultants on accessibility and usability from Fundosa Teleservicios
carried out a technical evaluation of accessibility based on a
twelve-point checklist synthesizing the levels of compliance with the
Web Content Accessibility Guidelines 1.0 (WCAG 1.0) proposed by
W3C/WAI. These twelve points mostly include priority 1 aspects from
the Guidelines, although some also come from other levels of priority.
The procedure for checking compliance required employing manual and
heuristic tests as the only valid approach to verify essential
qualitative features such as understandable links, or the correct use
of ALT labels and TITLE tags. Some tools such as the TAW Test for Web
Accessibility were expressly discarded due to their orientation toward
web page designers rather than for evaluating university portals and
other large corporate websites.
The web pages were visited using the most widely used graphic internet
browsers: Microsoft Internet Explorer 6.0, Netscape Navigator 7.0, and
Opera 7.2. Some features were deactivated during browsing in order to
check some of the points being tested, such as style sheets and
scripts.
The technical evaluation of accessibility sampled an average of six
web pages per portal. Pages were chosen to match the following
profile:
1.
Presentation, greeting, or language choice page (when available):
It may have hardly any content but it could pose a barrier to
accessing the site if it does not comply with accessibility
criteria.
2.
Home page: the portal’s most complex page, the way to get familiar
with the contents and a standard step to the other sections.
3.
Site map: to find your way around the site and access other
contents.
4.
Data table: an often-used way of laying out information.
5.
Forms: standard way to enter data for a variety of tasks (sending
suggestions, searches, even for course registration).
6.
Download page: a place students use to access contents, often
academic.
7.
Search results page: displays returns on searches for further
content.
The technical verification process was carried out during the period
from August 5 to August 30, 2004.
Fundosa Teleservicios concurs with the W3C/WAI recommendations that
the results of a technical evaluation of accessibility should be
complemented with an assessment based on feedback from users who have
different conditions of ability and disability. This approach is based
on the ISO 9241 quality standard, which defines usability as “the
effectiveness, efficiency, and satisfaction with which a product
allows specific users to reach specific objectives in a specific
context of use.”
User-centered assessment provides the means to check the “information
architecture,” i.e., how the information is organized (classified, and
labeling); how it is retrieved (navigation, search, and orientation
systems), and how it is structured for interaction (definition of the
interaction processes with the system).
The procedure used in this study is based on a User Feedback Test
styled as a self-administered questionnaire1 composed of a set of
tasks and subsequent questions to be filled in by the user following
prior instructions.
The sample of users who took part in the assessment survey was made up
of individuals with limited abilities of different kinds. They used
different kinds of technical aids for navigating and displayed
different levels of technical skill. Table 2 shows the profile of each
of the 6 users who took the assessment test.
Table 2- User profiles in the assessment test
Functional limitation
Technical aid used
Greatly reduced mobility of the hands
Trackball and head wand
Deafness
None used
Blindness
JAWS screen reader
Blindness
JAWS screen reader
Reduced eyesight
Screen resolution of 800x600
No notable limitation
None used
As in the technical evaluation, the results from the user feedback
questionnaires were collected, tabulated, and interpreted by experts
from Fundosa Teleservicios. The tabulation compares and weighs the
users’ feedback along with the objective results from the test tasks
on each portal by their efficiency, effectiveness, and satisfaction.
As an additional complement, the users were then organized into a
discussion group to go over their experiences. The tests were carried
out over the second week of September.
Analysis of the results of the technical evaluation of accessibility.
---------------------------------------------------------------------
The following results were obtained from the technical evaluation of
accessibility for each of the 12 criteria chosen for this study.
1. Validation of W3C technologies (priorities 1 and 2 in WCAG 1.0).
Both the HTML code and the Style Sheets (CSS) used on the pages should
be correctly expressed and validated by formal grammars, in this case
according to HTML and CSS specifications2. Any errors in the code make
rendering the page different depending on the browser used, since not
all browsers support all features.
The validation done for the study found code errors on all 84 pages in
the sample. Thus, none of them were in compliance of this criterion.
2. Frames (priorities 1 and 2 in WCAG 1.0).
It is not currently necessary to depend on frames to define the
structure of a website. If the frames option is chosen, their names
should be meaningful regarding their content, the relationship between
them should be clear, and an alternative should be given for users
whose browsers do not support frames.
Seven of the fifteen portals analyzed used frames. Three of them did
so constantly (UCM, USE, and UPV), and four made occasional use (CRUE,
UB, UAH, and UVA). None of them offered a title or description to
guide users. As a result, none of the pages complied with the
accessibility requirements for frames.
3. Forms (priorities 1 and 2 in WCAG 1.0).
The trouble some users may have when presented with forms includes not
knowing what data to enter or select in each field owing to the form’s
incorrect structure if the control tags are not arranged properly; not
being able to select and send data due to some browsers’
incompatibility with Javascript; not being able to tab through the
forms in the right order; or not finding the data arranged by topic or
concept on large forms.
None of the portals analyzed were found to comply with the
requirements for accessibility. The most frequently committed error
was the lack of correspondence in code between each tag and its
control, which caused great hardship especially for the people who
navigate with screen readers. Putting the tag on the left or above
each control makes rendering forms easier for tech-aided navigation.
This occurred on seven of the portals: MECD, UNED, UB, UPV, UV, ULPGC
and UN.
4. Text alternatives for multimedia items (priorities 1 and 2 in WCAG
1.0).
Some users can not see images (i.e., the blind, people using text-only
browsers, those who cancel downloads of pictures because their
internet connection is slow, etc.); others can not hear sound files
(the deaf, people who have no sound card on their PC, etc.). For them,
it is essential to provide some text-based alternative to multimedia
files.
The sample sites were checked to see if multimedia items also had text
alternatives for people who could not access them, and to verify that
the text alternatives found were adapted to the real needs of
different user profiles.
On all the portals in our study we found multimedia items without any
text alternative describing the image file. Nevertheless, the most
relevant images on some of the web pages on the UV, USE, and UGR sites
did have an alternative.
5. Headers (priority 2 in WCAG 1.0).
Headers (also known as “section titles”) are a fundamental way for a
website to mark the information structure on each page. Headers should
correctly indicate the level of depth: a level 1

header should
not be followed by a level 3

. Browsers such as Opera or JAWS
screen readers use headers to let users move around the page, a useful
feature for the blind and people with limited motor skills. This study
checked to see if headers were present, and if they were used
correctly.
None of the portals analyzed complied with this criterion for
accessibility. Most did not make use of header tags; they were only
found on the UV site, but even then they were used incorrectly (an H1
led to an H3, for example).
6. Relative units in Style Sheets (priorities 1 and 2 in WCAG 1.0).
Some people may need to be able to change the size of the text on
pages they visit in order to read the contents. Changing the font size
requires using either relative units or percentages (%), and
should be stated in the Style Sheets being used. Letter size, however,
cannot be changed if absolute units (in points, centimeters, or
pixels) are used in Style Sheets.
None of the pages in the sample used relative units in their style
sheet statements. Thus, none fulfilled this requirement, thereby
making it impossible for users with impaired vision or trouble moving
the cursor to change the size of the font.
7. Understandable links (priorities 1 and 2 in WCAG 1.0).
Links are an important structural item in a website, since they let
users navigate between pages and choose the content they want to
access. Texts and /or images used for a link should be
self-explanatory: they should give a clear indication of where they
lead to once clicked on. Some web browsers allow users to display only
text (or text alternatives to images); out of context, the link should
still make sense (for example, merely stating “click here” is
meaningless). The study also checked to see whether links to
downloading documents showed what format the target file is in.
Most of the links in the sample were indicative of their content.
Nevertheless, some pages were found to have image links with no text
alternative, while others had links with unhelpful text. Overall, this
requirement reached a considerable degree of compliance, especially on
the UV and USE portals.
8. Contrast (priority 2 for images in WCAG 1.0).
Some people can not view colors correctly, which makes it hard to read
the text on screen. For that reason, there should be enough contrast
between the text and the background color. In order to verify this
criterion, grayscale monitors were used to view the pages, and the
foreground image/background color contrast was scored (priority 2)
without quantifying the contrast between font color and background
(priority 3).
On the whole, the sample pages showed enough contrast between images
and background color, though the UOC and UNED portals were found to
have poor contrast between text and background color.
9. Semantic use of colors (priority 1 in WCAG 1.0).
Information conveyed by color on web pages should also be available
without color, as through context or markers. For the totally blind
and the colorblind, this accessibility requirement is vital. For other
people such as the learning disabled, the use of color can aid
navigation, but it should be done using the right context and markers
(for example, in style sheet statements).
None of the pages analyzed in our sample made use of color to convey
information. As a result, they all scored favorably in terms of
accessibility, since users unable to distinguish colors faced no
trouble in this respect.
10. Aligning content with tables (priority 2 in WCAG 1.0).
When tables are used for web page layout purposes (even though
nowadays we recommend using layers to lay out content), the content
should remain properly aligned when text-only browsers or screen
readers are used. Otherwise, serious problems may arise for rendering
the page contents.
Of the 15 portals reviewed, only the UOC portal had its contents
correctly laid out with tables. Errors in the remaining portals in
some cases surpassed the 50% mark.
11. Data tables (priority 1 in WCAG 1.0).
The blind or visually impaired may find it extremely complicated to
understand how data interrelates when put into tables with one or more
categories. Tables should always show headings for each column or row,
and use markers to associate header cells with their data cells in
tables involving two or more logical levels of headings.
We found some tables that did not display row and column headings
(which also affects their usability), and others that, while showing
headings for rows and columns, did not mark them as such by means of
code.
In the sample analyzed, no table was found to meet the required
criterion for accessibility.
12. Scripts (priority 1 in WCAG 1.0).
Care should be taken so that no functionality is lost when using
programming objects such as scripts should for any reason they not be
activated. Some browsers, such as Lynx, do not support scripts; others
may not have scripts activated because of the device being used or the
user’s lack of skill. In such cases, it is essential to offer an
alternative.
Of the portals analyzed, this requirement for accessibility was met
successfully by those of UCM and USE.
As an overview, Chart 1 shows how each portal in our study scored in
percent compliance with the 12 indicators used in the technical
evaluation of accessibility:
Chart 1
Classification of portals analyzed for accessibility, in percent
compliance

The most significant fact is that none of the university portals
managed to reach even a 50% compliance with the basic requirements for
accessibility sued in this study. In other words, none of them can be
considered to have passed the test. Clearly, there is considerable
work to be done.
The average compliance of the accessibility indicators is 32.11%, with
6 portals above average and 9 falling below.
When viewed in terms of the size of the university, schools with
higher enrollment fared better on accessibility. Four of them are
above the average, and the three others are close to it.
The three private universities are scattered in different positions in
the compliance ranking, which suggests no difference with publicly run
universities.
Both universities offering distance learning fall short of the
average. This fact is particularly relevant considering that students
use the web for online information much more often than in on-campus
educational settings.
The scores of the two general reference sites fall just below the
average for the sample, revealing a poor level of compliance (MECD at
33.3% and CRUE at 31.3%).
Of all the portals analyzed, the one attaining the highest score on
accessibility by the criteria used is the University of Valencia
General Studies (at 44.4%); at the other extreme, the University of
Alcalá de Henares comes in last place (at 21.3%).
Analysis of the results from the user feedback survey
-----------------------------------------------------
A particularly novel part of our study involved assessing the
accessibility and usability of the portals based on feedback from
users with and without disabilities. This section was carried out by
means of a task to be done by a variety of users with different ranges
of ability and functional limitations. The hands-on approach of having
users browse through and use a web portal helps identify any factors
they perceive as facilitating or hindering their use.
The users in our study carried out a series of five pre-designed tasks
per portal. After completing each task (or giving up for whatever
reason), each subject filled out a questionnaire consisting of 10
questions on a scale of 1 to 5 to indicate the user’s level of success
and satisfaction on performing each task. In addition to the user’s
satisfaction expressed on the questionnaire, the results were weighed
for efficiency and effectiveness in performing the tasks. A low score
(1) means the task was impossible to accomplish and/or the lowest
level of satisfaction. At the other end, a high score (5) indicates
the greatest degree of accomplishment and/or satisfaction. The middle
score (3) was used as the cut-off for a web page to pass the test.
Users were also encouraged to add their own comments on the
questionnaires to clarify any point. Finally, a discussion group was
held to go over their experiences together.
Table 3 shows what aspects were found to be helpful to navigation and
what were obstacles that hindered or completely impeded using the
websites. They are arranged according to the kind of limitation each
user had.
Table 3. Aids and barriers to navigation
IMPEDIMENTS
OBSTACLES
Aids
Visually Handicapped (the blind)
No text alternative to images directly related to performing a task.
Decorative images with no alternative text.
Correctly tagged links and images.
Mislabeled links for performing a specific task (i.e., mislabeled
search button).
Displaying search results with web strings from the website hosting
the search result. This creates confusion and does not give a clear
idea of what the website is about.
Correct verbalization of the controls for forms with a screen reader.
(Correct association of controls and labels)
Controls on forms are incorrectly verbalized by screen readers.
Incorrect association of text labels with controls (editing boxes,
drop-down lists, etc. in the form).
Search engine malfunction.
Providing accessible alternatives to otherwise inaccessible documents.
Using frames without title or content.
Not titling the search results page as such.
Specifying search results.
Use of non-accessible FLASH technology without an accessible
alternative.
Search engine results not done by section headers or titles to show
more accurate matches.
Providing an accessible map of the website.
Poorly structured information.
Relevant information with pre-assigned keyboard shortcuts.
Pages with automatic page refreshing, since screen readers reset the
cursor to the beginning of the refreshed page.
Using accessible Flash technology.
Opening new windows without warning users that they are on a new page.
Visually impaired
Errors upon loading the page (large image files).
Large numbers of pop-up windows on different topics.
Pop-up windows with additional information on the corresponding link.
Trouble viewing the page with high-resolution screens (e.g., 800x600).
Strange and oddly-placed menus.
Clear, concise information on the portal home page.
Very small menus and texts.
Choosing a different language does not display the same content.
Menus are always visible throughout the site.
Information is lost at higher screen resolutions.
Poorly indicated search bars.
Well-running search engine. Well-organized and structured results.
Non-functioning scroll bars, usually making the bottom of the page
unviewable.
Not being able to resize the page.
Hearing impaired
Text-heavy pages, too many links, and repeated use of unexplained
abbreviations to denote sections.
Images with no text alternative to show what the purpose of the image
is.
Simple, practical names for links.
Pop-up menus that hide the information underneath.
Search engines with few search options, making it hard to find
specific information.
Providing a site map or outline of the links.
Using complicated vocabulary or terms.
Outlines arranged in list form.
Not respecting the language chosen when entering the site despite
having selected it beforehand.
Intuitive pages with drop-down menus on the same page without adding
to download time.
Motor-skill impaired
Not being able to access all the content by keyboard input.
Drop-down menus requiring fine-honed precision to use them.
Fast, intuitive navigation.
Mislabeled links; links with confusing names.
Poorly indicated search engines.
Clear labels.
On the whole, the insights gleaned from the user feedback tests reveal
the troubles they had when trying to perform the assigned tasks. Some
of the problems were considerable obstacles—particularly for the
blind—and were only overcome by great effort and imagination, as was
the case when advancing through incorrectly labeled links. Still, the
overall assessment was positive for accessing most of the content
while highlighting aspects that can be improved or fixed for better
navigation.
The highest score users could give each portal was 50 points. The
total score obtained by adding up the 6 users scores (making 300
points the highest possible score) was re-scaled to 0 to 100 points,
and the theoretical cut-off set at 50.
Chart 2 offers the results obtained from the user feedback survey.
Chart 2
Global scores (adjusted) of portals in the User Feedback survey

On the whole, user assessment of the university portals was more
generous than the Fundosa Teleservicios experts’ technical evaluation
of accessibility. The average score on a scale of 0 to 100 was 61
points, and every portal except the University of the Basque Country
scored above the 50-point cut-off for passing. These more favorable
results from the feedback are deemed to owe greatly to the users’
adeptness and perseverance at navigating through websites fraught with
obstacles. This skill helped them carry out the tasks assigned, though
at the cost of having to invest greater time to do so. Another reason
for the comparative difference is that the low scores given by some
users were compensated by high scores given by others depending on
what kind of limitation each user faced. The score from the user
without any handicap or limitation did not substantially change the
overall score, even though his partial scores were on average slightly
higher than the scores from users with some kind of handicap or
limitation. In any case, from the users’ point of view, the web
portals of Spanish universities need considerable improvement if
certain groups of users with functional limitations are to be able to
access their contents.
The university portal receiving the highest user rating for
accessibility and usability was the UOC site, while the UPV took the
lowest score by a considerable margin (27 points).
The kind of problems users ran up against while carrying out the tasks
varied according to each user’s own functional limitation. Therefore,
a fully accessible and usable website needs to take all the criteria
into account if it hopes to reach all potential users.
Prior experience at navigating websites appears as the main factor for
a user to be able to access a website’s content, more than the
technical characteristics of the page itself. Even so, the users who
took part in our survey had to spend undue amounts of time to
accomplish the task, time they said they may otherwise not bother to
spend in a real situation. We have no reason to suppose that everyone
interested in accessing the content of a website has had enough prior
experience to dodge the difficulties revealed in this assessment.
Conclusions
-----------
The Spanish university web portals analyzed in this study showed a low
degree of accessibility in the technical evaluation carried out by
experts. None managed to achieve a 50% score in accessibility, and the
worst ones fell below the 25% mark.
Nevertheless, compared to earlier studies on university web portals
and accessibility, there has been some degree of improvement. In a
study by Egea (1998), none were accessible, and in one by Termens,
Ribera, and Sulé (2002) only 16 out of 256 pages sampled complied with
minimum requirements of accessibility.
In contrast, user feedback scores were higher, at an average of 61 out
of 100 points, with only one portal failing to make 50%. These higher
ratings should be viewed bearing in mind that it was likely the users’
own skill and perseverance, rather than the virtues of the technical
features of the page design, that helped them complete tasks they
might not otherwise do in the real world. Users pointed out a number
of obstacles preventing or hindering access to contents, and stressed
the need for the university portals under study here to improve their
accessibility.
The University of Valencia General Studies portal headed the ranking
on the technical evaluation of accessibility, with 44.44%. At the
bottom of the list was the University of Alcalá de Henares, at 21.31%,
a full 23 percentage points below the top.
The Open University of Catalonia, at 71.3 points, took top place on
the user feedback assessment, while the University of the Basque
Country came in last place, 27 points below, at 43.8 points.
If we convert the raw scores of the technical analysis to a scale of 0
to 100, and we add the points scored on the user feedback survey, we
find that the highest scoring portal belongs to the University of
Granada (108.18 points). Also above a hypothetical cut-off of 100
points are four other universities: Valladolid, Valencia, Seville, and
Open University of Catalonia. Last place, at 72.01 points, is the
University of the Basque Country.
Table 4 shows how each university portal is ranked by adding both
assessments.
Table 4. Final ranking.
UNIVERSITY PORTALS ANALYZED IN THE STUDY
Score out of 2001
1
University of Granada
108,18
2
University of Valladolid
106,51
3
University of Valencia General Studies
102,77
4
University of Seville
101,42
5
Open University of Catalonia
101,25
6
National Distance Learning University
96,19
7
Council of Rectors of Spanish Universities
95,00
8
University information from the Ministry of Education, Sports, and
Culture
94,58
9
University of Deusto
93,62
10
Complutense University of Madrid
90,35
11
University of Las Palmas G. Canaria
88,75
12
University of Navarra
83,26
13
University of Barcelona
80,65
14
University of Alcalá de Henares
76,31
15
University of the Basque Country
72,01
1 Scores on a scale of 0 to 200, the result of adding the adjusted raw
score from the technical evaluation of accessibility and the scores
given in the user feedback survey.
Our study reveals a number of unadvisable practices in the design of
contents for Spanish university web portals which hinder, and
occasionally prevent, people with limited functional abilities from
accessing the page. Similarly, several points were found largely to
comply with the criteria for accessibility on the aforementioned
portals.
The most relevant design errors (either because of their frequency or
the severity on accessing the information) were found to be the
following:
*
Forms that do not relate the tag to its control.
*
Lack or inadequate use of headers.
*
Scant use of text alternatives to images that convey pertinent
information.
*
None of the data tables on the portals analyzed complied with the
criteria for accessibility.
*
Seven portals used frames without adopting the criteria for
accessibility.
*
None of the pages had code that passed the W3C validation
technology.
*
Relative units on Style Sheets are not being used.
*
Programming objects such as scripts are used without providing
accessible alternatives.
*
A number of incomprehensible links, often for images without text
alternatives.
On the positive side, we can cite:
*
All the portals showed good compatibility with the three graphic
web browsers used in the study.
*
No semantic use of color was found on the portals analyzed.
*
Scores on contrast between background and images evaluated on
gray-scale screens were favorable.
*
Using tables for content layout surpassed the 50% mark on the
technical evaluation of accessibility.
Table 5. Internet addresses of the portals evaluated.
UNIVERSITY PORTALS
ANALYZED IN THE STUDY
INTERNET ADDRESS
Conferencia de Rectores de Universidades Españolas
http://www.crue.org
Información universitaria del Ministerio de Educación, Cultura y
Deporte
http://www.univ.mecd.es
Universidad Alcalá de Henares
http://www.uah.es
Universidad Complutense de Madrid
http://www.ucm.es
Universidad de Barcelona
http://www.ub.es
Universidad de Deusto
http://www.deusto.es
Universidad de Granada
http://www.ugr.es
Universidad de Navarra
http://www.unav.es
Universidad de Sevilla
http://www.us.es
Universidad de Valencia Estudios Generales
http://www.uv.es
Universidad de Valladolid
http://www.uva.es
Universidad Las Palmas G. Canaria
http://www.ulpgc.es
Universidad Nacional de Educación a Distancia
http://www.uned.es
Universidad País Vasco
http://www.ehu.es
Universitat Oberta Catalunya
http://www.uoc.edu
1 The questionnaire, filled out by users after doing 5 tasks on each
university portal, consisted of 10 questions per portal, in which the
user was to answer by rating each point on a scale of 1 (low) to 5
(high).
2 W3C provides both tools for validating code. For HTML, see
http://validator.w3.org. For CSS2, see
http://jigsaw.w3.org/css-validator/