AN INCLUSIVE APPROACH TO COOPERATIVE EVALUATION
OF WEB USER INTERFACES
Amanda Meincke Melo, M. Cecília C. Baranauskas
Institute of Computing, Unicamp, Caixa Postal 6176, Campinas, São Paulo, Brazil
Keywords: Accessibility, User Interface Evaluation, Cooperative Evaluation, Inclusive Design.
Abstract: Accessibility has been one of the major challenges for interface design of Web applications nowadays,
especially those involving e-government and e-learning. In this paper we present an inclusive and
participatory approach to the Cooperative Evaluation of user interfaces. It was carried out with an
interdisciplinary research group that aims to include students with disabilities in the university campus and
academic life. HCI specialists and non-specialists, with and without visual disabilities, participated as users
and observers during the evaluation of a Web site designed to be one of the communication channels
between the group and the University community. This paper shows the benefits and the challenges of
considering the differences among stakeholders in an inclusive and participatory approach, when designing
for accessibility within the Universal Design paradigm.
1 INTRODUCTION
Accessibility has been one of the major challenges
in user interface design for Web applications
nowadays. Besides enabling general access to
information needed for all citizens, it is a
requirement for the various domains such as e-
government and e-learning, in which knowledge and
education have been considered part of the mission
of nations and organisations.
Interface design for accessibility has long been
advocated as a fundamental requirement for
usability in general. Some efforts have also been
done towards the definition of recommendations for
providing the designers with tools to guide them in
designing and evaluating the applications
accessibility, especially Web sites.
Theories and methods in user interface design
have encouraged the participation of the user, in
different ways and through different phases of the
user interface production. The participation of users
in the interface design process has been considered
one of the best practices of the Human-Computer
Interaction (HCI) field. At the same time the
paradigm of Participatory Design (PD) has been
challenged when people with different types of
disabilities have been involved among the
participants of the design and evaluation process.
This paper brings this issue to discussion by
presenting a methodological proposal that extends
the Cooperative Evaluation Technique from PD with
artefacts of Organisational Semiotics (OS) to enable
an inclusive and participatory setting in a real
context of Web information system design. IPE is an
acronym for Inclusive Participatory Evaluation, a
new participatory technique we planned with the aim
of having people with different physical capabilities,
experiences and interaction styles participating
together in a cooperative evaluation of user
interface.
The IPE technique was applied successfully in
the context of “Todos Nós” project – an
interdisciplinary project being conducted in our
University, which aims at promoting educational
inclusion (Mantoan et al, 2005). Participants from
“Todos Nós” come from different professional
backgrounds, including people with disabilities. The
Web has been serving as an important
communication channel with people from inside and
outside our University, and Web-accessibility has
been one of the main concerns in the design of
“Todos Nós” portal.
Some results of using IPE are discussed in this
paper, especially considering the needs and benefits
of the technique in an inclusive environment to
evaluate accessibility and usability of Web user
interfaces. The paper is organised as follows:
Section 2 presents the theoretical background for
this work, which sets its basis on the concepts of
accessibility and the paradigm of Universal Design.
65
Meincke Melo A. and Cecília C. Baranauskas M. (2006).
AN INCLUSIVE APPROACH TO COOPERATIVE EVALUATION OF WEB USER INTERFACES.
In Proceedings of the Eighth International Conference on Enterprise Information Systems - HCI, pages 65-70
DOI: 10.5220/0002445500650070
Copyright
c
SciTePress
It is briefly presented here the two pillars of our
framework: Participatory Design and Organisational
Semiotics. Section 3 presents a summary of IPE
technique. Section 4 shows the first results of
applying it in a real context of user interface
evaluation. In Section 5 we conclude.
2 THEORETICAL
BACKGROUND
This work draws upon concepts and practices of
Participatory Design and Organisational Semiotics
to build a theoretical framework and understand
usability, accessibility and design for all as well.
In a broad sense, accessibility has been directly
related to the commitment of improving the quality
of life of elderly and people with disabilities (W3C,
2005; Bergman and Johnson, 1995). However,
taking into account the Universal Design philosophy
(Connell et al, 1997), it is possible to understand
accessibility as the easiness to approach and use
environments and products to the greatest extent
possible without discrimination. Although Universal
Design general principles point to ideal situations,
they constitute a valuable tool to guide the design
and the evaluation of more inclusive environments
and devices, which respect and consider the
differences among people.
Accessibility has been perceived as a necessary
attribute to the quality in use of software systems, or
to their usability (Bergman and Johnson, 1995;
Bevan, 1999; Graupp et al, 2003). If a user can’t
reach his/her objectives established in the interaction
with a computational system, the usability of this
system, relative to this user fails (Bergman and
Johnson, 1995; ISO, 1998). A design that
indiscriminately respects and considers the
differences among users must ensure that objectives
established in the interaction with a computational
system are reached (accessibility) with effectiveness,
efficiency and satisfaction (usability), to the greatest
extent possible (Graupp et al, 2003).
Besides Web-accessibility recommendations
(e.g. Section 508, Web Content Accessibility
Guidelines 1.0, and 2.0), in the Web context, there
are some techniques, which can be combined to
assess the accessibility of Web-based systems: the
use of different graphical and text-based Web
browsers, the use of assistive technologies,
automatic mark up languages validation,
accessibility verification with semi-automatic tools,
assessment based on checkpoints, evaluation with
users with different abilities and/or disabilities
(W3C, 2005; Graupp et al, 2003; Theofanos and
Redish, 2003). Nevertheless, there are still few
proposals in the literature considering the user’s
participation in an inclusive design setting.
Literature in Participatory Design has shown
different ways of including end-users in the process
of designing technology (Müller et al, 1997). PD
provides a set of techniques, which may support
different phases of the design lifecycle such as
problem identification and clarification,
requirements and analysis, high level design,
detailed design, evaluation, end-user customisation
and re-design.
In a PD perspective, a product is not only
designed for the users, but also with them,
collaboratively. In PD users’ engagement is
considered valuable to reach product quality, as it
allows a better understanding of their activities and
work context by the combination of different
experiences (Müller et al, 1997). At the same time,
PD can be useful to the users, inspiring them to think
about and analyse their own process of work. PD
could provide a valuable approach to inclusive
environments, where the individual differences must
be taken into consideration and users’ direct
involvement plays an essential role.
Particularly, the Cooperative Evaluation
Technique (Monk et al, 1993; Müller et al, 1997) is
a participatory practice to support the evaluation
phase, providing early feedback about re-designs in
a rapid iterative cycle. It can be used with an
existing product which will be improved or
extended, with an early partial prototype or
simulation, or with a full working prototype.
Designers without specialised knowledge of human
factors should be able to use it. Usually, an
evaluation team is formed of one end-user and one
developer to explore a software system or a
prototype, and develop criticism, so that changes
could be made to improve the product. In this work
we have adapted the Cooperative Evaluation
Technique with artefacts of Organisational
Semiotics to support participation of users with
different physical capabilities, experiences and
interaction styles in a cooperative and inclusive
evaluation of user interface.
Organisational Semiotics understands the
internal activities of an organisation, including its
information systems and its interactions with the
environment, as a semiotic system (Liu, 2000).
Organisation is understood in a broad sense,
meaning a group that shares some pattern of
behaviour and sign systems.
ICEIS 2006 - HUMAN-COMPUTER INTERACTION
66
We have been using some methods and tools
from OS to better understand the information system
behind the user interface in different levels (e.g.
physical, empirical, syntactic, semantic, pragmatic
and social). Particularly we have been using
methods from the set known as MEASUR in our
practice (Methods for Eliciting, Analysing and
Specifying User’s Requirements) (Liu, 2000) to
approach the design and the evaluation of technical
information systems, considering its social context
(e.g. responsible agents, behavioural patterns and
social norms).
Problem Articulation Method (PAM), for
example, is a method from MEASUR to be applied
in the initial phase of a project when problem
definitions are still vague and complex. Usually, it is
used to understand the aspects involved (e.g. needs,
intentions, existing conflicts, etc) in the design of an
information system, allowing a big picture of the
problem context, the main requirements and a shared
understanding among stakeholders (Liu, 2000). In
this work we adapted the Evaluation Frame – one of
PAM artefacts – to support the direct involvement of
users with disabilities in an inclusive participatory
evaluation of user interfaces.
3 IPE: INCLUSIVE
PARTICIPATORY
EVALUATION OF USER
INTERFACES
When designing with stakeholders from different
profiles (e.g. experiences, backgrounds,
capabilities), designers should be sensitive to
differences which come up and provide a flexible
setting to allow each stakeholder to participate
without discrimination.
Our first experience with IPE technique was
conducted with the participation of eleven members
of “Todos Nós” project. The participants have
different professional backgrounds and include
people with disabilities – one of them with low
vision and two with congenital blindness (both
Braille readers).
The technique was carried out to assess a
functional prototype of “Todos Nós” portal
(Mantoan et al, 2005). Its activities were planned to
take place during two to three hours. The aims of the
evaluation were to elicit accessibility and usability
problems, and to collect suggestions about the portal
interface design from prospective users.
During the exploration of “Todos Nós” portal, a
blind member, a low vision member and a sighted
member acted as users, while another blind member
acted as one of the observers. To guarantee the
participation of each member, some materials were
adapted to Braille (e.g. the participation term, the
task sheet, the observer's guide and the Evaluation
Frame) and/or printed with a larger font. Each
computer used during IPE activities had the
necessary technologies to support the interactions
between users and the portal (e.g. screen magnifiers
and screen readers). Although Perkins typewriters
were available for the blind participants, they
decided not to use them so their contributions could
be easily read by the other participants as well by
HCI specialists who would analyse them. Following
we summarise IPE technique.
3.1 Summary of the Technique
ABSTRACT Three to four teams are composed by
one end-user and two observers (one of them could
be the system designer or an HCI specialist). Each
team criticises the software system user interface or
prototype. After that they share their impressions
about the users’ experience, supported by an
evaluation frame adapted from the Organisational
Semiotics artefacts.
O
BJECT MODEL AND MATERIAL The software
system or prototype, a set of user’s tasks to help
focusing in the part of the user interface to be
evaluated, a set of questions – observer’s guide – to
help the observers to interact with the user,
recording materials (e.g. papers, pens and/or pencils,
audio and/or video records), a poster hanging on the
wall with the Evaluation Frame, and post-its to fill
the frame. Depending on the stakeholders’ physical
characteristics, it may be necessary to adapt some of
the materials and provide alternatives for note taking
activities.
P
ROCESS MODEL Starting Talk: The co-
ordinator explains the activities to be carried out, the
roles to be played by each participant, and the need
for agreement concerning ethical values. Phase 1
(Concurrent Cooperative Evaluation): three to four
teams are formed so they assess the software system
or prototype concurrently; (a) each team, composed
of a user and two observers, criticises the software
system user interface or prototype supported by the
user’s tasks and the observer’s guide. While one of
the observers keeps a dialog with the user during the
tasks performance, the other takes notes about this
dialog and the user’s interaction with the interface
(e.g. the user’s hypothesis, his/her choices, bad and
AN INCLUSIVE APPROACH TO COOPERATIVE EVALUATION OF WEB USER INTERFACES
67
good impressions, commentaries about the software
system or prototype); (b) each team talks about the
activity carried out, summarising good and bad
characteristics of the software system user interface
or prototype, as well as the user’s impressions about
the interaction activity itself. Phase 2 (Write-Paste):
all the teams share their impressions about the
software system user interface or prototype,
discussing issues/problems and solutions/ideas
regarding the user’s experience, writing them down
on post-its, and pasting these post-its on the
Evaluation Frame.
R
ESULTS Criticism of the prototype or software
system user interface, considering user’s experience,
especially as regards the accessibility; the register of
the problems found and some possible solutions,
taking into account the differences among
participants.
4 PRELIMINARY RESULTS
From the CONCURRENT COOPERATIVE
EVALUATION PHASE
we could perceive the users
had different strategies to browse, search and read
content in the portal.
In the execution of the first task regarding the
search for the Convention of Guatemala and its
interpretation, for example, the blind user, who
already knew the site structure by previous
experience on it, reached the “Law” section link
using TAB key and hearing screen reader feedback.
After entering in the “Law” section, she used the
screen reader search tool to look for the word
“decree” in the Web page. She perceived the second
occurrence of this word was part of the link to the
asked document, accessing it and completing the
task. This user spent about seven minutes to
complete the task, copying the answer from the Web
page and pasting it in a word processing document.
The low vision user, on the other hand, scanned
the navigation tree of the portal supported by the
mouse pointer and the screen magnifier – located at
the top of the screen, occupying a quarter of it. As
soon as this user found the “Law” section link, she
accessed it, selected the bold face presentation text
of the page with the mouse pointer and used the
Delta Talk software to help her reading the text. At
the same time she heard the selected text, she
scanned other parts of the Web page, but she didn’t
perceive there was a link to the asked document. So,
she decided to use de portal search engine to help
her in this task. The first attempt was unsuccessful,
as she entered an expression that referred to another
document. Entering a new keyword, she perceived
visual cues provided by the search engine, which
showed up the searched keyword in the page with
results, helping her to find the right link to the asked
document. After 35 minutes the execution of this
task was interrupted without conclusion.
The sighted user adopted an exploration strategy
to this same task. She accessed different sections of
the portal before entering the “Law” section. When
entering this section, she perceived the link to the
asked document, but understood it should be in the
“International Treats” subsection. This user spent
about seven minutes to complete the task,
summarising the answer in her task sheet.
The three users successfully completed the
second task, which asked for the definition of
“inclusion” in a published interview. The blind
initially tried to reach the interview through the
“Articles” subsection, without success. Thus, she
decided to take advantage of the portal search
engine, completing the task in five minutes. The low
vision user decided to make use of the portal search
engine at first, completing this task in seven
minutes. After exploring different sections, the
sighted user also decided to use of the portal search
engine, completing the task in nine minutes.
The blind and the sighted user also completed the
two extra-tasks, while the low vision user couldn’t
start them due to time constraints – IPE activities
should last no more than three hours. The first extra-
task asked for accessing a document in PDF format,
and the blind user reported some difficulties when
opening this kind of file through Web browsers as
some PDF files still have their content inaccessible.
The second extra-task, regarding the last published
news, was easily completed.
From the de-briefing – the talk established after
tasks execution – we emphasize the following
aspects:
The best features of the portal: for the
sighted user it was the possibility of having a broad
view of what could be found in the portal; for the
low vision user it was the yellow cues provided by
the search engine showing up the searched word or
expression; while for the blind user it was the
absence of Flash presentations, the long descriptions
provided for the links and the possibility of
accessing all the provided links.
The worst features of the portal: for the
sighted user it was the redundancy provided to the
main sections links (e.g. horizontal top menu and
vertical left navigation menu both providing access
to the main sections of the portal), and the use of the
same image to represent sections and subsections in
ICEIS 2006 - HUMAN-COMPUTER INTERACTION
68
the navigation structure; for the low vision user there
were nothing wrong with the portal; while for the
blind user it was also the redundancy considering the
main sections links: first the screen reader reads
aloud each section title and after it reads the long
description provided for each section link. For this
user each long description should be read together
with its link text. In fact, each link in the vertical left
navigation menu had a long description together
with its link text, but the screen reader didn’t give
her the chance to change between the link text and
the long link description. Figure 1 illustrates the
horizontal top menu and vertical left navigation
menu.
About the tasks: the three users reported the
asked tasks resemble activities they usually do in the
Web.
Analysing the Evaluation Frame data, from the
WRITE-PASTE PHASE, we identified some problems
without explicit solutions and vice-versa, suggesting
the existence of more problems and necessary
solutions than those explicitly pointed out by the
stakeholders. This is not surprising since stressing
all issues/problems and solutions/ideas weren’t the
only purpose of Write-Paste activity. This phase
aimed at allowing the group members to share their
experience with the portal, organizing and
registering their suggestions of improvement. The
following rates refer only to the issues/problems and
the solutions/ideas pointed out explicitly by the
group.
As it is illustrated in Figure 2, from 18
issues/problems related to interface design and
information design, 66,67% concern the blind user’s
experience, 61,11% concern the user with low
vision, and 66,67% the sighted user’s experience.
From the 21 reported solutions/ideas, 71,43%
concern the blind user, 71,43% concern the user
with low vision, and 66,67% concern sighted user.
Inspecting the Evaluation Frame we could perceive
many issues related to visual aesthetic (e.g. the need
for attractive visual elements, the need for better use
of blank spaces between groups of interface
elements), besides accessibility issues (e.g. the need
for better text description for the images, the benefits
of having access keys described together with their
link text, and the need for another colour schema to
cater for users with low vision). This could explain
the high percentage of issues/problems regarding the
sighted user besides the blind user. However, the
percentage of solutions/ideas related to the blind
user and the user with low vision is higher as a result
of the care to balance visual aesthetic proposals with
user interface accessibility.
If we consider only aspects related to
accessibility, we get a different picture from that on
Figure 2. As it is illustrated in Figure 3, the blind
user is the most affected by accessibility issues,
followed by the user with low vision. From 18
issues/problems 50% concern the blind user’s
experience, 38,89% concern the user with low
vision, and 27,78% the sighted user’s experience.
From the 21 reported solutions/ideas, 47,62%
concern the blind user, 38,10% concern the user
Figure 1: horizontal top menu and vertical left navigation
menu showing up the long description for its first link.
0,00%
25,00%
50,00%
75,00%
100,00%
Blind User
66,67% 71,43% 71,43%
Low Vision User
61,11% 71,43% 100,00%
Sighted User
66,67% 66,67% 100,00%
Issues/
Problems
Solutions/
Ideas
Functionality
Figure 2: Quantitative aspects from Evaluation Frame.
0,00%
25,00%
50,00%
75,00%
100,00%
Blind User
50,00% 47,62% 14,29%
Low Vision User
38,89% 38,10% 28,57%
Sighted User
27,78% 33,33% 28,57%
Is s ues /
Problems
Solutions
/Ideas
Functionality
Figure 3: Quantitative aspects from Evaluation Frame
regarding only accessibility problems.
AN INCLUSIVE APPROACH TO COOPERATIVE EVALUATION OF WEB USER INTERFACES
69
with low vision and 33,33% concern the sighted
user.
As there wasn’t a well-defined frontier between
issues/problems and solutions/ideas related to
functionality, we grouped them together in another
category. Among the seven suggestions regarding
functionality, only two of them are related to
accessibility: providing a form to send messages to
the portal team instead of only having the e-mail
contact published, and providing the users a way to
choose a different colour schema. The former would
benefit all users while the latter could improve the
interaction of sighted users and users with low
vision.
From this phase it is also registered that blind
participants want to have access to information
regarding visual aesthetic in images, but not only its
functional role. This wish was evident by the case of
the portal logo which functionally represents a link
to the portal main page.
5 DISCUSSION AND
CONCLUSION
This work has presented the Inclusive Participatory
Evaluation technique, which extends the
Cooperative Evaluation with the Evaluation Frame –
an artefact from OS. This technique supported the
assessment of a Web portal with prospective users in
an inclusive design setting.
Usually users from a Web application have
different backgrounds, experiences and capabilities.
The IPE technique was conceived to be applied in a
situation were users’ differences must be recognized
and considered in the design process.
The flexibility provided by the materials and the
behaviour of participants in the group dynamic
contributed to achieve results in which the solutions
were negotiated among people with different
capabilities and necessities in terms of user interface
interaction. This way IPE technique could allow a
designer to consider the real user’s experience (e.g.
technologies they use, the way the users deal with
their assistive technologies), and perceive the need
of balancing solutions that benefit their different
conditions.
While Concurrent Cooperative Evaluation
contributed to the exploration of a portal by different
users and observers showing them up some aspects
of interaction with the portal interface, the Write-
Paste activity helped them in organising and
registering their views and solutions, taking into
account the differences which exist among
themselves.
In summary IPE technique could support HCI
specialists and/or designer to assess technologies
with prospective users in inclusive design settings,
and effectively establish solutions committed to
different user’s needs. As a next step to this work,
we have been working on an Inclusive Web
Engineering Process that considers human factors
and users’ participation, in which this technique is
going to be integrated.
REFERENCES
Bergman, E., Johnson, E., 1995. Towards Accessible
Human-Computer Interaction. In: Nielsen, J. (ed.),
Advances in Human-Computer Interaction, Ablex
Publishing.
Bevan, N., 2001. Quality in use for all. In: Stephanidis, C.
(ed.), User Interfaces for All: Concepts, Methods, and
Tools, Lawrence Erlbaum.
Connell, B. R., Jones, M., Mace, R. et al, 1997. The
Principles of Universal Design, Version 2.0. Raleigh.
Retrieved April 2005, from The Center for Universal
Design, North Carolina State University:
http://www.design.ncsu.edu:8120/cud/
Graupp, H., Gladstone, K., Rundle, C., 2003.
Accessibility, Usability and Cognitive Considerations
in Evaluating Systems with Users who are Blind. In:
Stephanidis, C. (ed.), Universal Access in HCI:
Inclusive Design in Information Society, Vol. 4, Crete,
22-27, Lawrence Erlbaum Associates, pp. 1280-1284.
ISO, 1998. ISO 9241-11 Ergonomic requirements for
office work with visual display terminals – Part 11,
Guide on usability.
Liu, K., 2000. Semiotics in Information Systems
Engineering, Cambridge University Press, Cambridge,
218p.
Mantoan, M. T. E., Baranauskas, M. C. C, Melo, A. M. et
al, (n.d). Todos Nós. Retrieved July 2005, from State
University of Campinas network:
http://www.todosnos.unicamp.br/
Monk, A., Wright P., Haber, J., and Davenport L., 1993.
Apendix 1 – Cooperative Evaluation: A run-time
guide. In: Improving your human-computer interface:
a practical technique, Prentice-Hall.
Müller, M. J., Haslwanter, J.H., Dayton, T., 1997.
Participatory Practices in the Software Lifecycle. In:
Helander, Martin G.; Landauer, Thomas K.; Prabhu,
Prasad V. (eds.), Handbook of Human-Computer
Interaction, 2nd edition, Elsevier, 255-297.
Theofanos, M., Redish, J., 2003. Bridging the gap:
between accessibility and usability. In: Interactions,
vol. 10, issue 6, New York, ACM Press pp. 38-51.
W3C, 2005. Web Accessibility Initiative. Retrieved April
2005, from World Wide Web Consortium Web site:
http://www.w3.org/WAI/
ICEIS 2006 - HUMAN-COMPUTER INTERACTION
70