See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/226360940

Science Learning in Blind Children through Audio-Based Games

Chapter · October 2008

DOI: 10.1007/978-1-84800-136-7_7

CITATIONS READS

20 327

2 authors, including:

Jaime Sánchez
University of Chile
544 PUBLICATIONS 4,095 CITATIONS

SEE PROFILE

All content following this page was uploaded by Jaime Sánchez on 12 January 2017.

The user has requested enhancement of the downloaded file.

Science Learning in Blind Children through
Audio-Based Games

Jaime Sánchez, Miguel Elías

Department of Computer Science, University of Chile. Blanco Encalada 2120, Santiago,

Chile.
{jsanchez, melias}@dcc.uchile.cl

Abstract In this study we present AudioLink, an interactive audio-based virtual

environment for children with visual disabilities to support their learning of sci-
ence. AudioLink is a Role-Playing Game (RPG) for learning science concepts and
scientific reasoning through audio. We analyzed how blind learners can learn sci-
ence using audio as the main input/output interface and how to develop a chal-
lenging and engaging software. The usability of this software and a preliminary
study of the cognitive impact were also evaluated. Results indicated that users
considered the software was appealing, challenging and encouraging as a science
learning tool. AudioLink promoted a free and independent interaction at the users’
own paces. Evidence indicates that blind children developed scientific thinking
skills to identify a problem, build a strategy to solve it, and understand when it
was solved.

1 Introduction

Science learning is a process that uses mainly visual channels (charts, graphs,
simulations, to name a few), therefore the difficulties that a child may have when
learning science are stressed in children with visual impairments, since they have
greater impediments to access information, to learn basic mathematics and science
operations, and to solve problems [12, 0]. Computers are not just office tools to
work with; rather, they are intellectual tools to strengthen a series of cognitive
skills, and to solve problems creatively. Moreover, software –when properly util-
ized– can enhance learning. Another appealing way to support learning processes,
is combining games and education (known as edutainment), so children can en-
gage in learning situations playfully.
Most current software relies mainly on graphical interfaces, impeding the navi-
gation of users without the benefit of sight. Therefore, it is evident that contempo-
rary developments and approaches are unsuitable for visually impaired children
who already have problems to access information through traditional sources of

M. Redondo et al. (eds.), Engineering the User Interface,

DOI: 10.1007/978-1-84800-136-7_7, © Springer-Verlag London Limited 2009
88 Jaime Sánchez, Miguel Elías

interaction such as Braille books, relief maps and schemes, and audio-tapes. Also
there is a growing consensus that new designs and approaches should be consid-
ered [0].
A growing line of research in educational software for children with visual im-
pairments is using audio as the preferred sensorial channel when assisting the con-
struction knowledge and meaning making [0, 0, 0, 0]. These studies have shown
that audio-based interfaces can be used to promote learning and cognition in blind
children. They have also shown that virtual environments (represented through
audio) are a powerful incentive for blind children to develop and train cognitive
skills and to learn specific content. For instance, when virtual environments are
represented through 3D sound interfaces they tend to enhance a series of cognitive
processes [0] such as the development of general domain thinking skills such as
tempo-spatial orientation, abstract and short-term memory, and haptic perception
[0]. Although the literature describes software that supports the development of
mathematics learning and problem-solving skills with significant results for blind
children [0, 0], there is no relevant work in other science learning oriented soft-
ware (such as Biology, Physics or Chemistry). A few attempts have been made but
only with a rather limited user interaction, without providing interactive applica-
tions that encourage challenging and engaging them.
In this work we have designed, developed, and evaluated the usability of inter-
active audio-based multimedia software for learning science in blind children, us-
ing a gaming approach to enhance learning and cognition. We utilized a combina-
tion of incremental and evolutionary development, comprising the stages of
analysis, design, development, and validation. They considered activities that take
into account the particularities of developing educational software for blind chil-
dren. We evaluated the usability of this game using evaluation instruments for
end-users, experts and facilitators. A preliminary cognitive evaluation was also
conducted to verify whether or not users developed abilities to identify problems,
build strategies to solve them, and know when they have solved them.

2 Learning through games and simulations

Learning by using videogames is a new way of learning. Children enjoy the chal-
lenges and engagement of leaning with games. So, when having children as end-
users it is important to provide software that is game-oriented, so they can learn
playfully. As Kish et al. [0] stated one of the most important things when teaching
a child is that he or she is actually willing to learn. Providing an enjoyable envi-
ronment the learning experience can be most effective to construct knowledge and
be open to new stimulus and actions.
When combining education and entertainment (edutainment) it is possible to
provide a challenging and appealing learning experience, thus promoting effective
learning. Playing is an appropriate way of achieving a higher level of commitment
in learners as a consequence of the emotional attachment of the player with the
 Science Learning in Blind Children through Audio-Based Games 89

game. Shaftel et al. [0] stated that games have an important effect on the acquisi-
tion of problem-solving skills, allowing learners to try new strategies, stimulating
logical reasoning at the same time as also encouraging students to develop skills
such as attention to task, social skills –including taking turns and courtesy, coop-
eration with others, leadership skills, and fine motor skills. Since learning some-
times involves being wrong, games are an appropriate and appealing way of prac-
ticing what was learned over and over again, without serious consequences.
Furthermore, when a child finds something fun, it is likely that he or she repeats it
over and over again.
Currently, there is a clear trend towards using games and simulations for learn-
ing purposes, which has an important impact on how formal and informal educa-
tion can support each other to accelerate the learning process and the development
of high-order cognitive abilities, strengthening skill-based learning [0]. Navigating
a virtual representation of a real environment has an undeniable value and poten-
tial. Computer-generated simulations and virtual reality provide students with a
unique opportunity to experiment and explore a wide range of environments, ob-
jects and phenomena inside their classrooms or at home. They can observe and
manipulate objects, variables and processes in real time impossible to get by other
means. It is the adeptness of this type of technologies that makes them appropriate
instruments for the study of natural phenomena and abstract concepts [0]. They al-
low people to navigate a real environment that can be dangerous, hostile, or even
unreachable, and to simulate realities for safe learning.

2.1 The AudioLink software

AudioLink is a Role-Playing-Game where users control a main character to carry

out several quests or missions (Fig. 1). Each adventure is composed by sub-goals
of lower complexity necessary to fulfill a main goal. The player navigates through
connected scenarios that represent a virtual world with streets, houses, cities, and
the like, and interacts with elements and characters. This way, the player receives
information about the missions to be accomplished and the necessary clues. Each
quest involves rewards, such as new objects, access to other quests or areas in the
game, and so forth. Hence, it is possible to incorporate different science concepts
in each quest so that each adventure considers the learning of one or more associ-
ated concepts. In addition, the accomplishment of side-quests results in different
game endings. It is also possible to select between two levels of difficulty: normal
or easy.
AudioLink includes a tutorial that teaches basic game interactions (keyboard,
meaning of some audio cues, and others). The tutorial consists of a series of les-
sons grouped by the category to be learned –such as navigating and interacting
with the non-playing characters, or taking and using items, for example– and pro-
vides a space to practice what was learned independently at the user’s own pace.
90 Jaime Sánchez, Miguel Elías

Fig. 1. Interfaces of AudioLink

In this study we introduced a navigation model for virtual environment em-

ployed with great success in commercial videogames for sighted users that uses a
third-person aerial projection. This navigation model was adjusted to blind users
by adding 3D sound audio cues. AudioLink has also mini-games incorporated that
break the main metaphor temporarily aimed to develop additional skills (such as
improving coordination, spatial location, among others), and to become an addi-
tional incentive for software interaction. This leads to a more dynamic and attrac-
tive interaction, avoiding users to become tired and bored.
One of the mini-games developed is "Hammer Time" (Fig. 2-A), where the
player has to hit a worm that appears in eight possible directions (ahead, back,
right, left, and diagonals) and that quickly hides (besides of making fun of the
player), hence the user must press the numerical keypad in the right direction and
rapidly (before the worm hides). Another mini-game is the “Obstacle Race” (Fig.
2-B), where the metaphor shows a character running and avoiding obstacles
(through lateral movements and jumps), while at the same time trying to take the
emerging rewards.
As the embedded adventures are not trivial to solve (a sighted user that knows
the plot can end this game in no less than 40 minutes), game saving is provided, so
that players can keep their progresses and load them later. This function also
leaves an automatic registry of times, used objects, and other behaviors to make
further usability analyses such as how much time it takes users to complete a cer-
tain search and what zones were visited. To save the current game a menu is pro-
vided, which also lets users make volume adjustments (of the background music
and sound effects).

Fig. 2. The (A) Hammer-Time, and (B) Obstacle Race mini-games

 Science Learning in Blind Children through Audio-Based Games 91

Keyboard interaction was designed by considering keys already known by the

users who evaluated the software, from which we expect to make a standard in the
future. Blind users can detect the keys utilized rather easily by touching them and
because of their spatial location. The ‘F’ and ‘J’ keys are also detected by touch
since all keyboards have a notch in those keys, thus making references to the keys
located besides them. To navigate through the environment, arrow keys are used.
With the TAB key, the player can navigate through the equipment as in a circular
list. The player can ask for a description of the current equipped item (with the ‘D’
key), or try to use it (the ‘F’ key). To know more about the current scenario, a de-
scription can be asked (SPACEBAR), or the player can use the Zily character, a
fairy that follows the player, that can be sent to fly around the scene, indicating –
visually and through 3D sound– the position of objects and characters. Another
feat the player can take is through the action button (ENTER key), which activates
a contextual action, depending on the position of the player (open a door, take an
item, talk to a character). Finally, users can access a menu (‘J’ key) and save their
progress, load another game previously saved, adjust the volume, or end the game.

2.2 Representing the environment

The representation of the fictional environments of the game was developed

through a simple but highly representative model to retrieve and construct game
scenarios and different logical connections (or plot) from an external file. This
representation was described in a XML file with all attributes that characterize the
scenes. This allowed the modification of any element and even the storyboard of
the game just by editing the corresponding file without changing the source code.
Therefore, adding new stages or levels was possible by modifying only an external
file, with no changes whatsoever in the source code. From a different perspective,
by using this model it is possible to have infinite different games using the same
application. This can be done because the file, in addition to specifying how
scenes are constructed, defines logical connections that compose the game plot or
script, making it possible to modify the adventures of players through the external
file, and thus helping to overcome the lack of educational software for people with
visual impairments. AudioLink includes more than 150 different scenes to be ex-
plored. They are grouped in well-defined zones to promote long-term memory
use, since users have to remember global points to locate scenes inside a bigger
virtual world called the "Land of Imir" (Fig. 3). The software provides sequential,
parallel, optional, and alternative stories.
The abstraction representation used conceives the virtual world as a composi-
tion of different scenes. There are also groups of scenes called “Zones”, which re-
late the scenarios in a more abstract way, for instance, to associate different scenes
of a city or forest. The model starts from an abstraction layer that describes the
world as a combination of three main elements: Entities, Properties and Relation-
ships.
92 Jaime Sánchez, Miguel Elías

Entities are all elements present in an environment that can be seen as self-
contained pieces of data and that do not need to have a material existence. The
combination of one or more entities to each other assembles the representation of
the world. In a game environment, some entities could be the doors, the weapons
that a character can pick up, other non-playing characters to whom the player can
talk to, and the houses the player can enter, to name a few.
Entities can also be used to represent logical Relationships for embodying logi-
cal connections implicitly embedded in the represented world. For instance, in a
classic gaming situation, the main character tries to enter to secret area, thus he or
she has to pick up a certain key to unlock a secret door. In our model, there are
several entities (a key, door, and character) related to each other (the character can
grab the key and use it to open the door). Entities can also have a state to indicate
variable contextual information; the door can be opened or closed, a light switch
can be turned on or off, and so forth. Entities have different features and character-
istics. For instance, each entity has its own size, location and shape. Similar pat-
terns and features among related entities can be established, but the basic idea was
that there should not be two identical elements in the same place at the same time.
These characteristics that define the features of entities are denominated Proper-
ties.

Fig. 3. The “Land of Imir”, the virtual environment built

Each scene is made up of different elements which define objects to be created

in each scene, and the interaction with objects (elements to collide, to pick up, and
others). These elements are: back, facades, solids, characters, entrances, items,
doors, and points of use (Fig. 4).
Back corresponds to the background image that is used on the scene. It is the
element that has the greater depth in the z-axis according to the graphical rank of
layers and its function is just visual. Facades are surfaces placed over the back to
give additional information of the scene where the user is placed. A facade is a
visual aid, used to denote ways, footpaths, etc. But it also has a special sound as-
sociated. Thus a different sound is played if, for instance, the character is walking
on the grass, to the sound played if the character is walking on plain ground or a
wooden floor. Solids are all elements of the scene that the character can collide
with. These elements span from small fences to houses, and they are used mainly
to delimit the navigation through the scene. Characters are secondary characters
 Science Learning in Blind Children through Audio-Based Games 93

or non-playing-characters (NPC) that may appear in one scene and with whom the
user can interact. Entrances and Doors are elements of the scene used by the char-
acter to travel from one scene to another. They are invisible and facilitate leaving
to another scene when interacting with them. Items are objects of the world that
the player can pick up and use.
For the representation of logical connections the concept of Dependencies was
created. Dependencies are actions the user has to take to cause different associated
reactions, such as: picking up an object, using an object in a certain place (a point
of use), speaking with a particular character, and visiting a specific scene. The re-
actions can be new dialogues of characters, getting access to new zones and new
objects that can be taken.
Another crucial piece to achieve the script flexibility is the Points of Use, ele-
ments utilized to disrupt the logic of the game storyboard. A point of use is a spe-
cial zone within a scene where a particular item can be used indicating whether the
player loses or consumes that item. A point of use has an associated reaction that
consists of a particular audio cue played when using the item. It can be a dialogue,
an alert or anything that associates the fact of having used that item in such place.
A point of use does not limit the possibility of using objects; rather, it limits the
success or failure of using a specific object in a certain place. This is to say that
the logic behind using most objects consists of verifying whether the player is at a
point of use of the object to be used, so the corresponding reactions can occur.
There are also special items that can be used in any place (like the map or compass
items) and others whose logic is a bit more complex, such as the items used to ac-
cess to the mini-games.

Fig. 4. A model for representing the fictional environment

94 Jaime Sánchez, Miguel Elías

3 Methodology

A suitable model and methodology for designing and developing audio-based

games for blind children is proposed by [0]. The educational software model con-
siders the representation of the world, the knowledge to be learned, and the
learner. In this model (see Fig. 5), a combination of incremental and evolutionary
development is proposed considering stages of analysis, design, development, and
validation, stressing particular considerations when developing educational soft-
ware for blind children. This model considers a prototyping approach that allows
lowering the associated risk of failing to represent the user’s mental map adjusted
to potential unrevealed requirements at a low cost. A cycle on each stage was exe-
cuted producing a prototype. This process was repeated until a version that ful-
filled the requirements was produced. In these iterations the output from one cycle
was considered the input for the next one.

Fig. 5. Model of educational software for blind children

Analysis: In this stage we defined the set of tasks to be accomplished by the

student which lead to define software requirements (such as navigating, using ori-
entation resources, solving certain problems, and others). According to these re-
quirements different content was shown to the user, using a treasure hunt meta-
phor, where learners explored and interacted with the virtual world. Design: In
this stage the navigated virtual environment that users navigate was defined. The
game rules were settled, defining how the different metaphors (scripts) utilized
connected with each other. The world was designed as a set of scenarios grouped
in different zones with common characteristics (cities, forests, islands and others).
Development: This cycle consisted in the computer implementation of the models
that represented the fictional world and the user. We used an incremental model
which allowed the user to navigate some zones created in a coherent and simple
fashion when the first cycle ended. This way, on each cycle new zones, characters,
and functionalities were added to the game. The projections for different models
were designed and new information inputs and outputs flowed from and to the
user, mainly using 3D sounds. Validation: A usability evaluation was imple-
 Science Learning in Blind Children through Audio-Based Games 95

mented to obtain information about the user’s acceptance of the software, and the
match between his or her mental model and the representation included in the
software. Four instruments were considered for usability evaluation: end-users
questionnaire, heuristic questionnaire, direct observation sheets and the facilitator
evaluation questionnaire. Below we discuss the main results obtained in the us-
ability evaluation of the software. The data obtained was analyzed and studied to
figure out how the metaphors, models, and projections could be improved.

4 Usability Evaluation

Formally, the ISO organization proposed some usability definitions, depending on

the terms considered when specifying or evaluating usability [0], “The capability
of a software product to be understood, learned, used, and attractive to the user,
when used in a specified Context-Of-Use” (ISO/IEC 9126) and “The extent to
which a product can be used by specified users to achieve specified goals with ef-
fectiveness, efficiency, and satisfaction in a specified Context-Of-Use” (ISO
9241). The main importance of usability in software development is that it is a
critical factor for the system to reach its objective [0]. Users should have a real
feeling that the system will help them to achieve their tasks, avoiding becoming
reluctant to their use.
In spite of the known value of usability, there is still a trend in software engi-
neering to disregard the end-user, leaving to developers the self-testing of the in-
terfaces [0], making changes based on their own experience, mental model and
preferences. The problem with this is that developers cannot simulate the behavior
of blind end-users because their cognition and mental models are not alike. They
cannot assume their feelings and tastes when interacting with digital devices.
A usability evaluation was implemented in this study to ensure that the software
proposed was usable for blind children. End-user and heuristic evaluations were
implemented; concluding that the final software product developed can be utilized
by blind users independently. It was also verified that AudioLink is appealing, en-
couraging and challenging software that stimulates interaction.

5 Scenario and Sample

Software evaluations were carried out at the school for the blind “Santa Lucía”,
located in Santiago, Chile. Two special education teachers expert in vision disor-
ders also participated in the evaluations. The sample consisted of 20 students di-
vided into two groups: a group of eight children with ages between 8 and 12 years
old, half of them had low vision; and a second group conformed by 12 children,
with ages between 9 and 16 years old, where 9 of them had low vision.
96 Jaime Sánchez, Miguel Elías

All users were legally blind, which according to Chilean laws means that “an
eye is blind when its corrected visual accuracy is 1/10, or when the sight field is
reduced to 20 degrees”. This means that a legally blind person is able to see at one
meter (3.2 feet) what a sighted one can see at 10 meters (32 feet). They were all
familiar with interacting with a computer through a keyboard. All of them also had
expressive and comprehensive language skills and could utilize reading and writ-
ing systems (Braille or Macro-type), and had an IQ higher than 70. Some users
had some previous experience with game-oriented software; therefore they were
more critical when evaluating this type of software.

6 Instruments

Four instruments were used for the usability evaluation: 1. End-users evaluation,
2. Heuristic evaluation, 3. Facilitator evaluation, and 4. Direct observation. End-
user evaluation consisted in blind users answering a short usability questionnaire
elaborated by Sánchez [0] to evaluate the system usability considering interaction
attributes such as learning, efficiency, memory, errors and satisfaction. End-user
evaluations focused on opinions such as how users liked the software, what was
pleasant to them, what was not, and how useful the software was to them. The
heuristic evaluation [0, 0] is a methodology to find usability issues in the design
of interfaces, and involved expert evaluators that judged the interface considering
a series of validated usability principles –the heuristics– that are general rules that
describe usable interfaces (Table 1). Heuristic evaluation was implemented
through an extended questionnaire with a likert-type scale elaborated by Sánchez
[0]. Direct observation was applied to get to understand the fluidity of particular
situations in a fast and direct way. The facilitator evaluation was designed consid-
ering facilitators as end-users when facilitating the interaction of children with the
software. Apart from their own interaction with AudioLink, facilitators were pre-
sent during the interaction of children with the software, acquiring knowledge and
unique points of view that shaped them when evaluating this software usability.

Table 1. Usability Heuristics.

Number Heuristic
1 Visibility of system status
2 Match between the system and real world
3 User control and freedom
4 Consistency and standards
5 Error prevention
6 Recognition rather than recall
7 Flexibility and efficiency of use
8 Aesthetic and minimalist design
 Science Learning in Blind Children through Audio-Based Games 97

9 Recognize, diagnose, and recover from errors

10 Help and documentation
11 Content design
12 Velocity and media

7 Procedure

End-users evaluation considered the interaction of blind children with a series of

incremental prototypes. Users interacted with the interfaces independently and
could ask facilitators for assistance. When the interaction session was over, a us-
ability questionnaire was applied. Questions were read and explained to children
by the facilitators who also wrote down their answers. Six evaluations with end-
users were implemented, from September to November (2005), each of them of
approximately one hour of duration.
Three evaluators –experts in usability with experience in working with soft-
ware for visually impaired people– participated in the heuristic evaluation. During
the evaluation session, each expert inspected the interface independently, judging
each component against the heuristics. This procedure provided a list of usability
issues, classified by heuristic (visibility, flexibility, and so on). Then, we con-
ducted dialogue rounds in order to understand the problems faced by the experts
and to discuss possible redesigns. The heuristic evaluation was taken in one ses-
sion of 40 minutes for each usability evaluator.
The facilitator evaluation was taken by two special education teachers, expert
in vision disorders. They had not interacted with the software before the evalua-
tion session –only observed the children’s interactions. The facilitators interacted
with the software and then evaluated it by answering the facilitator questionnaire
independently. The evaluation was taken in one session of 30 minutes.
In all of these evaluations an observer was present without interacting with the
participants, but recording the usability problems detected, and taking a photo-
graphic register of the evaluation sessions. This information that was later ana-
lyzed and considered for interface redesign.

8 Results

The first significant result of the usability study was that in all evaluations a very
high rate of acceptance was obtained. End-users evaluation showed that the navi-
gation model and the proposed interaction were appropriate and highly accepted
by users, who were able to carry out all interaction features embedded in the soft-
ware. Fig. 6 shows the end-user evaluation of the first prototype (first test) and the
98 Jaime Sánchez, Miguel Elías

final product (last test) for children with low vision (left) and totally blind children
(right).
Users liked the software during all implementation stages and always found it
entertaining. The final product was more challenging, simple to use, and highly
encouraging. End-users always stated that they would play again the software, re-
flecting that the interaction was attractive and pleasant to them. The importance of
usability evaluations is reflected in the affirmation about the software being easier
to use in its final version rather than in the first prototype. This happened not only
because the level of appropriation of users was higher, but also because the soft-
ware was adjusted after each evaluation session, reflecting more accurately the
needs and interests of end-users through incremental prototypes.

Fig. 6. Results of usability evaluations by children with residual vision.

Software audio cues and sounds were also modified in order to achieve a higher
quality and making them more representative of the audio stimuli and more attrac-
tive for children. This influenced users by assuming a significantly higher score in
the affirmation about how they liked the sounds of the software, when comparing
prototype and final product evaluations. Thus, according to the evaluation taken to
end-users with low vision, we can conclude that the interface is highly interactive,
attractive, and appropriate to them.
Even though during the blind children’s evaluation it was possible to observe a
small decrease in the final score of some statements –when comparing the first
and last evaluation– in average a favorable score still remains. The main score de-
crease was in the statement “the software is easy to use” because a virtual envi-
ronment of higher extension and more a complex storyboard was obtained in the
final product, so users had to interact with a higher number of elements and a
greater number of characters. This imposed greater cognitive load to users, be-
cause it was designed with the purpose of obtaining a more challenging and en-
couraging game. Another collateral effect of increasing the complexity of the
game was observed in the decrease of the score for the statement “I would play
again with the software”. This seems to be somewhat contradictory when consid-
ering the increase to a perfect score in statements that evaluate this software as en-
tertaining, challenging and encouraging. It can be explained because by making
the virtual environment more complex, which implies a higher audio memory ca-
pacity of users, leading to a higher cognitive loading, and thus decreasing the
 Science Learning in Blind Children through Audio-Based Games 99

scores about whether they would play again the software. Also, they did not use
the game saving functionality, so each evaluation began from the same starting
point, thus losing interest in the game. In spite of these slight decreases we can
conclude that considering the high scores obtained, the proposed interfaces are ap-
propriate and were highly accepted by these users.
The tutorial embedded in the software showed to be useful and provided an
added value to the game by teaching users to learn the basic interactions inde-
pendently at their own pace. Evaluations showed the appropriateness of using
quadraphonic sound for the mental construction of the space by blind users (the
first prototype had only stereo sound). The appropriate and correct design of the
interfaces favored visually impaired users to play and achieve goals effectively
and efficiently, gaining satisfaction during interaction. These characteristics al-
lowed adding a greater complexity to the virtual environment by adding more mis-
sions, characters, and elements. Users also showed acceptance and encouragement
to interact with the mini-games presented. Although each prototype evaluated had
a greater complexity than the previous one, they also demonstrated in each inter-
action session a greater degree of software appropriation.
The heuristic evaluation also provided positive results for software improve-
ment. Some software errors were detected thanks to this evaluation and fixed for
the final version. This evaluation also led us to find some problems with icons and
navigational issues, and even to suggest new functionalities for the software. The
facilitator evaluation also indicated that this software presents and integrates sci-
ence contents clearly. This makes AudioLink an appropriate tool for children to
learn curriculum contents and to develop and exercise problem-solving skills. Fa-
cilitators also mentioned that this software had a very usable interface for low vi-
sion users, encouraging blind users to use it.

8.1 Cognitive Evaluation

As part of this study, a preliminary cognitive evaluation was implemented to

measure the impact of using AudioLink on the development of scientific thinking
skills. Two instruments were used: 1. Logging the actions taken by users during
playing –which were recorded automatically by the computer– and 2. An open
question questionnaire. We first analyzed the actions that are implicit part of the
development of scientific thinking skills, such as observing, identifying problems,
looking for solutions and evaluating. We considered variables such as interaction
time, rate of visited zones, and number of items that were found and used. We
concluded that users were able to perform most interactions, they highly motivated
themselves, and they developed scientific thinking skills. Likewise, the ways hy-
permedia stories are presented encouraged users to interact with the software in
the long run, since there are several alternative endings depending on the user de-
cisions in the game.
100 Jaime Sánchez, Miguel Elías

Additionally, an open question questionnaire –with a problem-solving orienta-

tion– was applied. Users were asked to indicate what problems they faced in the
game, how they knew when something was a problem, and how they solved it. Fi-
nally, they were asked to make an evaluation of their own interaction. Results
showed that users identified problems, when mentioning, for instance, that "we
must do things that are not said"; meaning that there were things not explicitly
stated in the game that they had to infer. Users also understood that information
was distributed in different zones and to gather it they had to navigate the virtual
environment. They also recognized that they enjoyed and solved different problem
situations after planning and executing diverse strategies.
Last year, we implemented an in-depth study to evaluate the impact of working
with AudioLink and cognitive tasks on science learning by blind children, in a six-
month field study [0]. This study showed us that blind children were able to learn
and practice scientific method processes and enjoyed learning new topics by wid-
ening and enhancing their theoretical conceptualizations and acoustic perception.
There was a manifest physics learning gain and analytical cognitive skills were
enhanced. In some participants these gains meant significant achievement. Chil-
dren were also able to map, use and understand a game with a complex dynamic
and interaction, and solved complicated science problems. We concluded that the
use of AudioLink, combined with cognitive activities, promoted scientific content
learning and also enhanced problem-solving skills during experimentation con-
veyed through the application of the scientific enquiry method.

9 Conclusions and Future Work

In this study we have designed, developed, and evaluated the usability of Audio-
Link, interactive audio-based multimedia software for visually impaired children
for the learning of science. A usable product was developed for the user’s inde-
pendent utilization. We verified that the software was appealing, encouraging, en-
gaging, and challenging. In addition, users rated the usability of AudioLink
highly, demonstrating that user-centered design favors the final interaction of end-
users with the software product.
This software has an enormous potential by incorporating new scenes and ad-
ventures from an external file. This makes it possible to think that in the near fu-
ture the teachers of blind children, their parents and even themselves, could be
able to create their own games and extend existing ones. Modifying the game by
means of an external file (for instance, using the game editor currently being de-
veloped), can partly help to overcome the lack of educational software for people
with visual impairments.
We validated that users developed abilities and strategies to identify a problem,
solve it, and know when it is solved. We also observed that users spontaneously
contrasted information they cooperatively exchanged to help each other. Children
were able to play and understand a game with increasing complexity. They were
 Science Learning in Blind Children through Audio-Based Games 101

able to solve problems of high difficulty and to explore a world of a higher magni-
tude than anything they may had experienced previously.
When creating a generic RPG engine for blind users, this software can be ap-
plied for learning concepts of any given subject. This game genre has shown to be
very appropriate, useful and appealing for most children that played it. It was also
very important to experiment with a widely used game model developed for
sighted users and successfully implemented in commercial console games. This
demonstrated that perhaps it is not necessary to redesign the whole model to make
this type of software available for visually impaired users. Rather, it is possible to
consider some adjustments, oriented to support users in their own particular needs.
Finally, the role that AudioLink can play is in the same direction of those tools
that help to provide opportunities to children with visual impairments to be more
integrated and included in the society.

References

1. Bevan, N. (2001). International standards for HCI and usability. In: International Journal
of Human-Computer Studies, Vol.55, No. 4, Oct 2001, pp. 533-552.
2. de Freitas, S., Oliver, M. (2006). How can exploratory learning with games and simula-
tions within the curriculum be most effectively evaluated? Computers and Education 46
(3), pp. 249-264.
3. Eriksson, Y., & Gärdenfors, D. (2004). Computer games for children with visual im-
pairments. Proceedings of The 5th International Conference on Disability, Virtual Real-
ity and Associated Technologies, 20-22 September 2004, New College, Oxford, UK,
pp. 79-86.
4. Fowler, M. (1998). Mocks objects aren't stubs, Testing Methods: The Ugly Duckling,
Keeping Software Soft. Retrieved January 4, 2007 from
http://www.martinfowler.com/distributedComputing/duckling.pdf
5. Kish, D., Bleier H., Moser S. (1998). Facilitating Movement and Navigation in Blind
PreSchoolers: A Positive, Practical Approach. Online report. Retrieved Jan. 13, 2007
from http://www.waftb.org/Facilitating%20Movement.pdf.
6. Kurniawan, S. H., Sporka A., Nemec, V. and Slavik, P. (2004). Design and user evalua-
tion of a spatial audio system for blind users. Proceedings of The 5th ICDVRAT, 20-22
September, 2004, New College, Oxford, UK, pp. 175-182.
7. Mastropieri, M., Scruggs, T. (1992). Science for students with disabilities. Review of
Educational Research, Vol. 62, No. 4 (Winter, 1992), pp. 377-411.
8. Molich, R., Nielsen, J. (1990). Improving a human-computer dialogue, Communications
of the ACM 33, 3 (March), pp. 338-348.
9. Nielsen, J. (1993). Usability Engineering. Academic Press Professional, Boston, MA.
10. Nielsen, J. (1994). Heuristic evaluation. In Nielsen, J., and Mack, R.L. (Eds.), Usability
Inspection Methods. John Wiley & Sons, New York
11. Panchanathan, S., Black Jr, J. A., Tripathi, P., and Kahol, K. (2003). “Cognitive Multi-
media Computing”. Proceedings of IEEE International Symposium on Information Sci-
ence and Electrical Engineering (ISEE 2003), November, Fukuoka, Japan, pp. 13-15.
12. Sánchez, J. (2001). Visible Learning, Invisible Technology. Santiago: Dolmen Editions.
102 Jaime Sánchez, Miguel Elías

13. Sánchez, J., Baloian, N. (2006). Modeling 3D interactive environments for learners with
visual disabilities. In K. Miesenberger et al. (Eds.): ICCHP 2006, LNCS 4061, pp. 1326
– 1333, 2006. Springer-Verlag Berlin Heidelberg 2006.
14. Sánchez, J., Elías, M. (2007). Science Learning by Blind Children through Audio-Based
Interactive Software. Annual Review of CyberTherapy and Telemedicine (in press).
15. Sánchez, J., Flores, H. (2004). Memory enhancement through audio. Proceedings of
ASSETS 2004, Atlanta, Giorgia, USA, October 18-20, pp. 24-31.
16. Sánchez, J., Flores, H. (2006). AudioMedia: Multimedia for blind people. 11th Annual
CyberTherapy 2006 Conference: Virtual Healing: Designing Reality. June 13 - 15,
2006. Gatineau, Canada, pp. 65
17. Sánchez, J., Sáenz, M. (2005). Developing Mathematics Skills through Audio Inter-
faces. Proceedings of HCI 2005. Las Vegas, Nevada, USA, July 22-27, 2005.
18. Sánchez, J., Sáenz, M. (2006). 3D sound interactive environments for blind children
problem solving skills. Behaviour & Information Technology, Vol. 25, No. 4, July -
August 2006, pp. 367 – 378
19. Shaftel, J., Pass, L., Schnabel, S. (2005). Math Games for Adolescents. TEACHING
Exceptional Children, 37(3), pp. 25-30
20. Strangman, N., Hall, T. (2003). Virtual reality/simulations. Wakefield, MA: National
Center on Accessing the General Curriculum. Retrieved 4/3/2007 from
http://www.cast.org/publications/ncac/ncac_vr.html.