1 Introduction

flameluxuriantData Management

Dec 16, 2012 (4 years and 6 months ago)

219 views



detectLD:
D
etect
ing

University Students with L
earning
D
isabilities

in
Reading and W
riting in
the
Spanish

L
anguage



Carolina Mejía, Jonathan Clara, Ramón Fabregat,

Institute of Informatics and Applications, University of Girona,

17070 Girona, Spain

{carol
ina, jclara}@eia.udg.edu, ramon.fabregat@udg.edu


Abstract
:
This
work

is
focused
on
the

inclusion of university student
s

with learning
disabilities

in
a learning m
anage
ment s
ystem (LMS)
.
W
e have developed
a
self
-
questionnaire

to detect
the
student populati
on
that

presents or could present a learning
disability

in reading
and writing
in the Spanish language
.
T
his work will
contribute to
the design and development
of a
software tool

(
detectLD
)

that support
s

this
self
-
questionnaire

and the creation of a
nother
self
-
questionnaire

that can be
used for the detection of different learning problems.
detectLD

was tested and evaluated
by teachers
,

responsible for administering it
,

with a
group of
student
s

(age
d

20

30 years)
at

the University of Girona.



1
Introducti
on


Since the
19
80s, the
fields

of psychology and pedagogy
have

made

important contributions
to

understanding
some students’
learning difficulties

(learning
disabilities
)

in

reading and writing.
I
n recent years
,
however,

there
has been

pa
rticular concern
a
mong
the
teachers
who are
review
ing

their teaching practices to
improve
the processes
involved in reading and writing
and to learn

how to assess and
intervene in these
difficulties
.
Several studies
have explored
learning
disabilities

in reading and writing

in children
:

identifying
the
population of children with learning disabilities, evaluating cognitive processes
involved
determining

their
specific deficits
,

and creating

intervention programs to i
mprove
the learning
deficits presented
(Nicolson, 1990;
Met
sala, 1999;
Guzman, 2004)
. M
any of those programs
have been
supported
by

information a
nd
communication technologies (e
.
g.
,

software)
that
tend to increase

student
’s

motivation and
personaliz
e

the
learning process

(Barker, 1995;
Wise, 2000; Macaruso, 2008;
Rojas, 2008)
.

In this research we
address another population:
university students.
R
esearch

in related works

has
shown that
learning
disabilities

persist into adulthood
(Finucci, 1986;
Bruck, 1993
;
Booth
, 2000; Wilson, 2001
)
.
For this reason, it
becomes
ne
cessary to study the cognitive processes that can be altered in university students,
and how
their
deficits can be treated.

W
e are
initially
interested in
detecting

the university
students

that
have

or could
have
a learning
disability

in reading and/or wri
ting (
e.g.
,

dyslexia)
. I
n Spa
i
n
, university students

are not
asked

for this
information when entering the university, and
therefore

the number of
specific cases

in university classrooms

is
unknown
.

I
n this work we in
troduce
detectLD
;
a software tool

that
t
akes advantage of
web
-
based technologies to
detect

specific reading and writing
problems
, to

inquire about the
school life and family background
, to know
which habits

student
s

have in reading and writing, and to identify students with an early diagnosis of

the
problem and whether they are
in treatment.

We
design
ed

detectLD

based

on the work of
Giménez
(2010)
, who built
a
self
-
questionnaire

that was
completed

manually by students at the University of Malaga (Spain), making
it
possible to detect dyslexia,
dys
graphia, dysorthography, and other
learning
deficits

among
the student population
.
That

self
-
questionnaire

was
adjusted in accordance with the objectives of this research
(
to detect potential learning
disabilities

in
reading and writing

in university stude
nts
)
; the initial

questions

were

improved and

rewritten
to adapt them
to
a
web
-
based scenario
.

Based on
the adaptation of the
self
-
questionnair
e

in this research
,

we implemented and tested
detectLD

as a generic
software
tool
to detect learning
disabilities

in reading and writing
, which

could
also
be integrated in
a
learning management system (
LMS
)

and it allows
others
types of
self
-
questionnaires

to
be
embedded

for
detect
ing

other

learning problems (e.g.
,

dyscalculia,
dysphasia,
and
a
ttention deficit disord
er
).

Furthermore,
we developed
a case study with a pilot group of students
(age
d

20

30 years)
from the
University of Girona to test the functionality and
the
usability
of
detectLD
,

to
determine how easy it is to read


the

self
-
questionnaire

by the students,

and to calculate the average time that the students take to complete it
.
T
he
case study
helped us
appreciat
e

the
effectiveness of the
self
-
questionnaire

and led to
suggestion
s

made
and
improvements recommended by the teachers who were responsible for
givi
ng
the
self
-
questionnaire

to the
students
.

This

paper is structured as follows. I
n the second section
we explain our
reasons for

studying

university
students and
build
ing

a web
-
based

application.
In the third section
we describe
the
conceptual model,
the
a
rchitecture, and
the
modules

of
detectLD
.

In the fourth section the implementation and testing of
detectLD

are

presented
.
The fifth section describes a case study with a group of students
from

the University of Girona.
Finally
, the sixth section
draw
s

some

conclusion
s and
presents
proposals for future work
.



2
Motivation


2.1
Why university students?


Until r
ecently, learning disabilities (LD) had been studied
very little
at the university level (Gregg,
2007; Sparks, 2009; Jiménez, 2004). Today it is a

topic of interest
because of
the high prevalence found in this
population. According to the British Dyslexia Association

(BDA, 2010)
,

it
is estimated that between 10% and
1
5
% of the student population
worldwide
have some LD, while

in Spain, according to

J
aen Dyslexia Association

(
ASDIJA
,
2009), it is estimated that the prevalence
at

university is
between

6% and 8%, although an exact
percentage is unknown
and i
t is believed that this
percentage
may increase in coming years. Because the
percentage of the
pop
ulation of university students with LD is high
,

services for these students

must be
increased
, and resources that treat the specific deficiencies
of
the student
s

must be created
to improve their
academic performance
so
that they
can
advance at

the same pac
e as their peers. Moreover, these services and
resources
may

motivate
otherwise reluctant
students to
register in

different university programs
.

M
any of them
do
not register becau
se of their LD (
Ingesson, 2007)
,

which
make
s

them
lose their self
-
esteem and
feel
intimidated and
unable
to continue beyond high school
.


2.2
Why a web application?


T
he Internet is a technology that facilitates communication and
student
accessibility
so
we design
ed

and implement
ed

the software tool

as a web application
that do
es

not
need to be installed or upgraded and that
is
used more and more as
web
-
based technologies
constantly
develop.
These advantages provide better
ways
for
the community of students, teachers and
others such as psychologist, pedagogue or counselor
to inter
act
.
Moreover, there is no technological discrimination because
detectLD

runs on a web
browser such as Mozilla
Firefox

or
Internet Explorer
and
can
be used with
any
operating system

(Linux, Windows, Mac, Solaris, etc.).
Furthermore,
because
detectLD

is
a w
eb application,
it can be integrated
directly
in Moodle
or
using web
services

in others

LMS
.



3
detectLD:
Learning Disabilities

Detection
Software Tool


The main objective of
detectLD

is to enable the creation, delivery, and review of the results of
the

self
-
questionnaire

to detect learning
disabilities

in reading and writing in university students. Therefore,
here we
define
some functional and nonfunctional
requirements
of
it
, and
specify
its behavior using UML
:



Roles:

Since
detectLD

is a
web applicatio
n
that
creates

self
-
questionnaires

to check
for
possible learning
disabilities

in the university context, we can define three types of users:
Experts
,

or
user
s

responsible for
performing tasks related to
creating
self
-
questionnaires

and checking the result
s;
Teachers
,

or
u
sers
responsible for scheduling and activat
ing

the
self
-
questionnaire in their

class
es and
chec
k
ing

overall results
of the course;

and
Students
,

or
u
ser
s
who

respond
to
the

self
-
questionnaire

activate
d

by the teachers
.




Platform:
W
e used

o
pen source technology
: the
Apache

web server
(Apache, 2010)
which
has support for
PHP

scripting language

(PHP, 2010)
and
the
relational database

management

Postgres

(Postgres, 2010)
,
all
installed on a Linux
o
perating
s
ystem
server
.





Use cases:
In order to
specify and detail the behavior of
detectLD
,
we defined
some use cases for the
software tool
.
They
have been
organized
into different
functional
groups for better interpretation, as shown
in the use
case
diagram in Figure 1
.



Figure

1
.

Use case
diagram
.


The architecture of the
software
tool is designed to facilitate
interaction

between the different modules
in relation to the user who uses them.
For each type of
user

a different interface is presented depending on the
permissions and tasks that can be de
veloped. In the architecture we specify the behavior of
the tool
,
summarizing both
the

components and the relationships. In Figure 2 the components
compris
ing

the architecture
are show
n
: 1) the
expert module
, 2) the
teacher module
, 3) the
student module
, 4
) a
web server

that supp
orts the
tool and 5) a
database
.



Figure 2
.

DetectLD
a
rchitecture
.



A
s
can be
seen in
the use case diagram in Figure 1,
detectLD

has
three

main modules
with
functions

in
accordance with the tasks to be developed for each
previous
ly defined
role
.
T
hese modules are described

as
follows
:



Expert module
: This module was designed and implemented for the exclusive use of a subject matter
expert (psychologist, pedagogue, or counselor). The module allows the creation of different self
-


ques
tionnaires needed to detect any
type of
learning

problem

(in our case learning
disabilities

in reading
and writing). According to the use cases
presented
for the experts
in Figure 1
,

e
xperts

can create differ
ent
self
-
questionnaires, divide

them

into
sectio
ns or issues to
be
evaluate
d
, create different types of questions
(
y
es/
n
o, single choice, multiple choice and
open
-
ended
), make changes or deletions,
and
check (
consult

and analyze)

the results of
the
self
-
questionnaires.
Figure 3 show
s

the initial interf
ace of the expert
module.



Figure 3.

Expert

module

initial interface.




Teacher module
: This module was designed and implemented for the teacher or tutor who can see the
different self
-
questionnaires
as
well as view the overall results of the
course.
Also
,
the
teacher

is

responsible for activating the self
-
questionnaires to be
completed
by the students. Figure 4

show
s

the
initial

interface of the
teacher module
.



Figure 4
.

Teacher

module

initial interface.




Student module
: This module is exclusively
used

to
complete

the self
-
questionnaires
that
have

been
previously activated by the

teacher. Initially this module asks the student
s

some general

information
(academic program, date of birth and sex) and then
it
presents the interface to
complete

the self
-
ques
tionnaire. Figure
5

shows the
initial

interface of this

module, and Figure 6

shows the interface to start
filling
in
the self
-
questionnaire
.




Figure 5
.

Student
module
initial interface.



Figure 6
.

Interface of the self
-
questionnaire.



4
detectLD
Impl
ementation



detectLD

was
implemented

with standard technology
and
considering characteristics of reusability,
interoperability, accessibility, and extensibility, so that
it
was easy to integrate into the structure of an
educational system.

In our work a

p
articular case was built with the patterns of
Moodle

styles
so that later it
could be
integrated with
the
LMS.


This implementation was
based on a combination of
open source technology
such as Linux

Operating
System
(specifically Ubuntu 9.04, with kernel v
ersion 2.6.28)
, Apache

Web Server (version 2.0)
, Postgres

(version 8.3.9)

and PHP

(version
5.2.6
)

aforementioned
, due to
their

great
popularity
on

different servers,
their

high performance,
and their
easy setup and acquisition. In t
erms of security, passwo
rds use

the widely used


reduction algorithm MD5, and permission levels
are
set by the developer according to the modul
e that the user
wishes to enter
.

detectLD
was

also

implemented using
standard
programming technologies
that
provide
interoperability

with
other web
-
based systems.



Interface based on XHTML and CSS
.
The main
programming technologies

considered for the
development of the user interface are: 1)
XHTML markup language
with
content
-
oriented structure

for
the
presentation of documents via the Web, a
nd 2) CSS
to create
style

sheets for the presentation of
content
regardless

of the structure of the page content (so we can separate presentation and content

layers
).



PHP and JavaScript
.
PHP is used to manage the site structure and dynamics of the applicat
ion
and
generate

greater interactivity in the interface
(
part of the password encrypting and dynamic presentation of
the contents in function of the data stored in the database
,

such as different questions and results).
JavaScript was used mostly for
the
v
alidation of forms and
for
part of the password encrypting
.



Asynchronous communication with the server via XML
.
AJAX technology enables communication
with the server
with

asynchronous JavaScript and XML
and
using the XMLHttpRequest object
. This
technology

was used to increase interactivity with the user interface, allowing the partial updat
ing

of
website
s

without having to reload the entire page
(i.e.
,

reloading some forms
according to

what the user has
chosen in other forms
on
the same page, without refre
shing it)
.

Considering the aforementioned technologies,
detectLD

could be integrated in Moodle
,

a web
-
based
LMS that supports
the management

of
different

activities.
Figure 7
shows the interface of the
detectLD

in
Moodle
.



Figure 7
.

Expert’s u
ser interfa
ce in
Moodle
.


Moodle

is a
LMS

with
great
pedagogical and technological flexibility
and
usability, and with the
support of a large community of users around the world.
It

has been
developed as an
open
source
educational
application with a free software lic
ense
, and is currently the LMS used at the University of Girona.
D
uring the
development of this work we
configured
detectLD

to make
it work as a module of Moodle (developed with
Moodle version 1.9.10),
so that
it can be used by different university teacher
s interested in detecting students
who may have a learning
disability

in reading and
/or

writing. To
integrate
detectLD

with Moodle, we took
advantage of our
software
tool
, which

was

implemented based on an architecture similar to Moodle (Linux,


Apache, P
HP

and Postgres), so it was easy to
make minor changes to
integrate folders and da
ta storage
within its
structure
.

Finally, during the implementation phase of the
detectLD
, we
tested
the different modules separately.
These
tests

revealed the need for
changes

in interface design and programming
to
achieve a better system
performance. The types of
tests

used were
:
connection to the database, requirements, inspection
software/programming and functional testing of the different parts

(such as creating/deleting ne
w
tests
,
adding/changing sections or adding/changing/deleting questions)
.

To
test

the real
-
time performance and usability

(Sauro, 2005)

of
detectLD
,

a case study

was designed
with a
pilot
group of students from the University of Girona. This
case

study is
presented in the next section.



5
Case Study
:
Performance

and Usability


Eight

students from the University of Girona

participate
d in the

case study
. For this sample
both male
and female
students from different academic programs and levels, aged betwee
n 20 and 30,
were
selected
.
W
hether or not the student had an LD

was not taken into account
, because the aim of this
case study

was to
assess the
performance

and usability of

detecLD
,

how well participants
underst
ood

the questions
in the
self
-
questionnaire

and how long they could take to complete the self
-
questionnaire
.

During the
case study
, students were accompanied by an examiner (university
teacher
)
experienced in
managing

detectLD

and responsible for taking note of possible questions and problems
of
th
e student
s

while
they are
using the
software
tool

an
d
filling

in

the self
-
questionnaire
.

When student
s

have

completed
the
self
-
ques
tionnaire, the examiner give
s

them

a survey to fill
in
by hand
and
intended to
evaluate the
performance
and
usability
of the

software tool
and
whether or not
they understood

the questions
in the self
-
questionnaire
. The
survey questions are
presented
in Table 2
.


Two examiners conducted

the
case study
. The first examiner was responsible for applying the self
-
questionnaire to unde
rgraduate students, while the second examiner was in charge of graduate students.
Examiners recorded the time each student took to fill
in
the self
-
questionnaire
.

They found

that students could
complete it
in
8
to
12 minutes,
a relatively short time
.
T
he
s
tudents’
answer
s

showed
that
none of the eight
who
filled
in
the self
-
questionnaire
had
a

potential
LD
,

which means
that

the tim
e taken to complete the form by
students

with
an
LD

could be
longer
. Examiners also noted that
detectLD

was very
user friendly

a
nd intuitive
:

the student
s

never had questions about how to access and complete the self
-
questionnaire. Finally, the examiners
reported
that students only had
difficulty with specific questions
,

which were subsequently reviewed and
restructur
ed by the expe
rts
.

The survey
used to gather
s
tudent comments consisted of
seven
evaluation
questions
.

T
he student
s

chose
the
most appropriate

response

on a scale of 0
-
4
based
on
their
perception
.

In addition
, at the end of the
survey
a space was
left

where the student
s

could include more comments if
they
wish
ed
.

The

survey
’s results
showed
a good level of
performance and
usability of the system as well as
a
good understanding of the questions
in the
self
-
questionnaire
.
The results
obtained
of the questions
are presented

in Table
1
.


Evaluation questions

No




Much

Satisfaction

0

1

2

3

4

Do
you
think the
self
-
questionnaire

seem
ed

easy

to fill in
?



1


7

93
.
75%

Do you think the
self
-
questionnaire t
ook
short

time

to fill

in
?


1


2

5

84
.
38%

Do

you
think

the questions
were easy to understand
?




1

7

96.88%

D
o

you think the font
size of the questions
was

appropriate
d
?


1

2

1

4

75%

Did
the topic of the self
-
questionnaire
makes

you feel
motivated
to fill
it in
?


1


2

5

84
.
38%

Do you think the focus o
f
the
questionnaire (learning
disabilities

detection)

is important?



2

1

5

84
.
38%

Would
you recommend
that
a friend fill it

in
?

1


1

2

4

75
%

Table

1
.

Results of the survey filled
in
by students.


With the survey responses we calculated the satisfaction
percentage of the students about each
question. The formula to calculate

this satisfaction percentage was
:




% Satisfaction = 100 * (
R
1*0 +
R
2*1 +
R
3*2 +
R
4*3 +
R
5*4) / (
N
*4)


In the formula

R1, R2, R3, R4 and R5 represent the 5 possible responses that can

be given to each
question

(
scale of

0

to
4
),
N

is the number of stude
nts who responded to the survey, and the multiplying
numbers
are the values
assigned to each type of response

(scale of 0

to
4).
These results showed that the
satisfaction
percentage o
f student
s is

quite high in terms

of performance and usability of

detectLD
.

In addition to the results obtained in the above table, we also analyzed each of the comments the
student
s included

on the survey answer sheet
.

In general the results were
very pos
itive, except for some
comments on the wording of questions; the student
s

did not understand some of them

well
.

Finally,
some
students expressed interest
in knowing the results of
the
self
-
questionnaire

and the steps

taken

to follow it up
.

T
he results obta
ined in this case study
will
help us
improve

the
software tool

and the self
-
questionnaire
in order to use it with
more

groups of university students.



7 Conclusions and Future Work


A web
-
based software tool
to

detect university students with learning
d
isabilities

in reading and writing

called
detectLD

was designed and developed
.
Moreover, the software tool is extensible to others learning
problems (e.g., dyscalculia, dysphasia, and
attention deficit disorder
).
Th
is

software
tool

has been tested
with

a
g
roup of
eight
students from
the University of Girona
in
different academic programs and
at different
levels.

The platform chosen for development

open source and providing flexibility

was adequate
despite
the limitations of this type of software. The
standa
rds
-
based implementation

allowed the creation of a
better
-
designed
web
site to produce
clear

code and
be
compatible with other systems.

W
e
also
found that the
implemented

software tool

is usable and
it has a good
performance
because
the
results of the test
s

with student
s

were
very
satisfactory.

Some problems related with the wording of questions
were displayed by students, but the experts have already solved.

As a future work
we

are going to deliver the self
-
assessment
to

students of different academic
prog
rams, in order to get a percentage of students that can
present a

learning disability in reading and/or writing
at the University of Girona.

W
e
propose
to
link

detectLD

with a
n

assessment battery
for
cognitive process
es

for
university
students (Mejia, 2010
),

which will be integrated into the
Moodle

platform. This
integration
will allow
all
the
teachers of the university community to
access
detectLD
, and therefore
to know
of
and learn
about
possible cognitive

deficits
of
their
students
.



References


Barker,

T. A. &

Torgesen, J. K. (1995).
An evaluation of computer
-
assisted instruction in phonological
awareness with below average readers
.

Journal of Educational Computing Research
, 13,
pp.
89
-
103.


Booth, J. R., Perfetti, C. A., Mac Whinney, B.
&

Hunt, S. B. (
2000).

The association of rapid temporal
perception with orthographic and phonological processing in reading impaired children and adults
.

Scientific
Studies of Reading
. 4, p
p
. 101

132.


Bruck, M. (1993).

Word recognition and component phonological process
ing skills of adults with childhood
diagnosis of dyslexia
.
Developmental Review
, 13,
pp.
258
-
268.


Finucci
,
J
.
M
., Gottfredson,
L
.
S
., & Childs,
B
.
(1986).

A follow
-
up study of dyslexic boys
.
Annals of Dyslexia
,
35,
pp.
117
-
13
6.


Giménez de la Peña, A., Bu
iza, J.J., Luque, J.L., & López, J. (2010).
Cuestionario de Autoinforme de
Trastornos Lectores para Adultos (ATLAS)
.

XXVII Congreso de la Asociación Española de Logopedia,
Foniatría y Audiología (AELFA 2010)
, Valladolid (
Spain
).




Gregg, N. (2007).
Underse
rved and Unprepared: Postsecondary Learning Disabilities
.
Learning Disabilities
Research and Practic
e
, 22,
pp.
219

228.


Guzmán, R., Jiménez, J. E., Ortiz, M. R., Hernández
-
Valle I., et al. (2004).
Evaluación de la velocidad de
nombrar en las dificultades
de aprendizaje de lectura
.
Psicothema
, 16,
pp.
442
-
447.


Ingesson, G. (2007).

Growing up with Dyslexia: Interviews with Teenagers and Young Adults
.
School
Psychology Inter
national
, SAGE Publications, 28
(5),
pp.
574
-
591.


Jaen Dyslexia Association


ASDIJA
(2009). Unidad De Atención A Estudiantes Con Dificultades Especiales.
[Online at http://www.dislexiajaen.es/]. Accessed 10/12/09.


Jiménez J
.
E
.
, Gregg N,

&

Diaz M. A
.

(2004)
.
Evaluación de habilidades fonológicas y ortográficas en
adolescentes con dislexia

y adolescentes buenos lectores
.
Infancia y Aprendizaje
,
27(1), pp.
63
-
84.


Macaruso, P.
&

Walker, A.

(2008).

The efficacy of computer assisted instruction for advancing literacy skills in
kindergarten c
hildren
.
Reading Psychology
, 29
(3), p
p
. 266
-
287.


Mej
ía, C., Fabregat, R., & Marzo, J. L. (2010
).
Including Student's Learning Difficulties in the User Model of a
Learning Management System.
XXXVI Conferencia Latinoamericana de Informática

(CLEI 2010), Paraguay.


Metsala, J. L. (1999).
The development of pho
nemic awareness in reading disabled children
.
Applied
Psycholinguistics
, 20,
pp.
149
-
158.


Nicolson, R. I. &

Fawcett, A. (1990).
Automaticity: A new framework for dyslexia research
?
.

Cognition
, 35,
pp.
159
-
182.


PostgreSQL Global Development Group (2010).

PotsgreSQL.
[Online at http://www.postgresql.org/].
Accessed
20/07/10.


Rojas R. E. & Jiménez J. E. (2008).
Diseño y validación de un videojuego para el tratamiento de la dislexia
.
PhD the
sis, University of La Laguna, p. 606.


Sparks, R
.

L. &

Lovett
, B. J
. (2009).
College students with learning disability diagnoses: Who are they and how
do they perform
?
.

Journal of learning disabilities
, 42 (6), pp. 494
-
510.


Sauro, J. & Kindlund, E. (2005).
A Method to Standardize Usability Metrics Into a Single Score
.
C
onference on
Human Factors in Computing Systems (CHI 2005)
. Portland, Oregon
(
USA
)
.


The Apache Software Foundation (2010).
Apache HTTP Server.
[Online at http://www.apache.org/]. Accessed
20/07/10.


The British Dyslexia As
sociation


BDA (2010),

Dyslexia
and Specific Learning Difficulties in Adults. [Online
at
http://www.bdadyslexia.org.uk/
]. Accessed 22/02/10.


The PHP Group (2010). PHP. [Online at
http://www.php.net/
]. Accessed 20/07/10.


Wilson, A.
&

Lesaux, N. (2001):
Persistence of phonological proces
sing deficits in college students with
dyslexia who have age
-
appropriate reading skills
.

Journal of Learning Disabilities
, 34(5),
pp.
394

400.


Wise, B. W., Ring, J. & Olson, R. K. (2000):
Individual differences in gains from computer
-
assisted remedial
rea
ding
.

Journal of Experimental Child Psychology
, 77 (3), p
p
. 197
-
235.





Acknowledgments


The a
uthors would like to thank the Spanish Ministry
of
Science and
Innovation

for financial support
through
the

A2UN@
project

(
TIN2008
-
06862
-
C04
-
02/TSI
).
T
hanks to the

Scholarship
Program

of the University of
Girona
,
reference

BR08/09. Thanks also to Almudena Giménez and Alicia Diaz for their collaboration and
comments.