A Game to Target the Spelling of German Children with Dyslexia

Abstract

Playing error-based exercises presented in a computer game was found to significantly improve the spelling skills of chil­ dren with dyslexia in Spanish. Since there are no similar error-based exercises for German, we adapted the method to German and created 2,500 new word exercises. Since dyslexia manifestations are language dependent, the replica­ tion of the method required (i) collecting new texts written by German children with dyslexia; (ii) the annotation and the linguistic analysis of the errors; and (iii) the creation of exercises as well as their integration in the tool.

Full text

A Game to Target the Spelling of German Children
with Dyslexia
Maria Rauschenberger1,2
Silke Fuechsel1
1University of Applied Sciences Emden/Leer
2OFFIS, Institute for Information Technology
Oldenburg, Germany
rauschenberger.maria,sfuechsel
@student.fh-oow.de
ABSTRACT
Playing error-based exercises presented in a computer game
was found to significantly improve the spelling skills of chil-
dren with dyslexia in Spanish. Since there are no similar
error-based exercises for German, we adapted the method
to German and created 2,500 new word exercises. Since
dyslexia manifestations are language dependent, the replica-
tion of the method required (i) collecting new texts written
by German children with dyslexia; (ii) the annotation and
the linguistic analysis of the errors; and (iii) the creation of
exercises as well as their integration in the tool.
Keywords
Dyslexia; Serious Game; Spelling; Written Errors; German.
Categories and Subject Descriptors
K.4.2 [Computers and Society]: Social Issues—Assistive
technologies for persons with disabilities; K.3 [Computers
in Education]: Computer Uses in Education—Computer-
assisted instruction.
1. INTRODUCTION
Motivation. In Germany, between 5 and 12% of stu-
dents have dyslexia, a neurological learning disability which
impairs a person’s ability to read and write. Also in Ger-
many, only 25% of the poor spellers achieve average spelling
performance during primary school [7]. In a longitudinal
study Esser et al. [1], showed that children who were di-
agnosed with dyslexia at the age of 8 achieved lower school
performance and higher rates of unemployment than a con-
trol group (measured afterwards at the age of 25). Later,
Schulte-K¨
orne et al. [8] showed that it takes longer for a
child with dyslexia to achieve school grades that are com-
parable to the others, even if they have an above-average
socio-economic status and are given effective treatment.
Permission to make digital or hard copies of part or all of this work for personal or
classroom use is granted without fee provided that copies are not made or distributed
for profit or commercial advantage and that copies bear this notice and the full cita-
tion on the first page. Copyrights for third-party components of this work must be
honored. For all other uses, contact the Owner/Author(s). Copyright is held by the
owner/author(s).
ASSETS’15, October 26–28, 2015, Lisbon, Portugal.
ACM 978-1-4503-3400-6/15/10.
DOI: http://dx.doi.org/10.1145/2700648.2811345.
Luz Rello
Clara Bayarri
Azuki Gòrriz
Cookie Cloud
Barcelona, Spain
luz,clara,azuki@cookie-cloud.com
Background Research. To support the spelling of chil-
dren with dyslexia in Spanish, Rello et al. [6] designed a
method to create exercises based on the linguistic analysis
of the errors made by children with dyslexia. It is called
an error-based method because it uses written errors as a
starting point since (1) dyslexic writing errors are related to
the types of difficulties that they have [10]; and (2) readers
with dyslexia cannot consciously see errors in words [3]. The
exercises were integrated in an iOS game, Dyseggxia (Pirule-
tras) [5], which was subsequently evaluated in a school. Over
eight weeks, the authors carried out a within-subject exper-
iment with 48 children from 6 to 11 years old. The children
who played DysEggxia for four weeks in a row had signifi-
cantly fewer writing errors in the tests than those playing
Word Search –the control condition– for the same time [6].
Solution. In this demo we present the German extension
of that tool and explain the adaptation of the method to
create word exercises for children with dyslexia. It is worth
noting that despite the universal neuro-cognitive basis of
dyslexia, dyslexia manifestations are variable depending on
the degree of consistency and regularity of the language or-
thography. German has a semi opaque language, that is, the
relationships between letters and sounds are not as regular
as in other transparent languages as Spanish [9]. In addi-
tion, it also has more complex syllabic structure [9]. Hence,
the adaptation of the method from Spanish was not straight-
forward, raising a number of challenges. To the best of our
knowledge, this is the first approach to create German error-
based exercises for people with dyslexia.
2. RELATED WORK
Specifically for German we found Lernserver1, a tool to
diagnose and provide support exercises. The tool was evalu-
ated in 2008 with the Landesinstitut f¨
ur Schule, where 78.2%
of the students (N = 3,798) improved their writing by using
this tool [2]. Klex 11 2 is a tool to practice vocabulary in
German adapted to a certain school degree. The exercises
of Klex are not error-based since only the correct words are
presented. We could not find literature on how Klex was de-
veloped and how the exercises have been created. Similarly,
CESAR schreiben 2.0 3 is a strategy and educational game,
which includes listening, spelling and vocabulary exercises.
1http://www.lernserver.de/home.html
2http://www.legasthenie-software.de
3http://www.ces-verlag.de/deutsch/Schreiben2_0/
schreiben2_produktinfo.php5

Figure 1: Screenshots for two types of exercises: cut
into words (left) and change a letter (right).
To the best of our knowledge, there is no tool to support the
spelling performance of children with dyslexia for German
based on written errors that they make.
3. GERMAN ADAPTATIONS
Below, we explain the steps to replicate the method and
describe the specificities added for German.
Texts Collection. The first step was the collection of
dyslexic errors. We collected 47 texts (homework exercises,
dictations and school essays) written by students with ages
ranging from 8 to 17 years old. A total of 32 texts were
written by children who have been diagnosed with dyslexia.
The remaining 15 texts correspond to students with a high
spelling error rate that were identified by their teachers.
The students attended either primary school, comprehen-
sive school (Gesamtschule), high school (Gymnasium) or a
special school (F¨
orderschule).
Error Annotation. We manually extracted the errors
and annotated them with phonetic and visual features [4].
These features were specific to the German writing system
and phonetics. For instance, German has letters and sym-
bols, that do not exist in Spanish or English, e.i. ß, ¨e, oa, ¨ ¨
and ¨
u. In addition to that, the handwriting systems broadly
used in schools (Lateinische Ausgangsschrift) presents dif-
ferent letter forms, so the visual analysis of the letters was
language specific.
Linguistic Analysis. Third, we analysed the errors lin-
guistically. These analyses had to be specifically performed
for German specific syllabic structure and semi opaque or-
thography, for instance, the group of letter <sch> are pro-
nounced as only on phone [S]. This resulted in different error
categories as well as different linguistic criteria.
Creating the Exercises. We created exercises taking
into account the previous linguistic analyses. As a result we
had 8 types of exercise: (a) add a letter; (b) remove a letter,
(c) change a letter (see Figure 1, right), (d) choose the cor-
rect word ending, (e) cut into words (see Figure 1, left), (f)
order the letters or syllables, (g) capital letter, and (h) wrong
capital letter. The last two exercises are specific to German
where nouns are always written with capital letters, and
missing capitalization is a frequent misspelling. Also, cut
into words exercises included words and morphemes specific
to German such as the trennbare verben, that is, verbs is
that are preceded by prefixes.
Difficulty Levels: Thre are five difficulty levels: Initial,
Easy, Medium, Hard, and Expert. Higher levels included
less frequent words, longer words, and words with a more
complex morphology. The criteria to define the levels were
adapted to German. For instance, all levels presented longer
words and morphology played a stronger role because Ger-
man has a more complex morphology and longer words.
4. CONCLUSIONS AND FUTURE WORK
Using the adaptation of the method described above, we
manually created 2,500 German exercises. The exercises
were integrated into the iOS app Dyseggxia (Piruletras).
4
For future work we plan to carry out a longitudinal eval-
uation of this tool in a German school. Moreover, we will
be able to compare how error-based games are more or less
effective for certain types of languages.
Acknowledgements
We deeply thank J¨
org Thomaschewski for making the con-
nections between teams so this collaboration was possible.
We also thank Miguel Ballesteros for his help creating the
German dictionaries; J¨
org Thomaschewski and Hendrik Witzel
for helping with the phonetic error annotation; and Erin
Brady for her helpful comments reviewing the text.
5. REFERENCES
[1] G. Esser, A. Wyschkon, and M. Schmidt. Was wird
aus achtj¨
ahrigen mit einer lese- und
rechtschreibst¨
orung? ergebnisse im alter von 25
jahren. Zeitschrift f¨
ur Klinische Psychologie und
Psychotherapie, 4:235–242, 2002.
[2] L. Flug. Projekt-evaluation bremen: Erprobungsphase
des lernservers an bremischen schulen, 11.2007.
[3] L. Rello. DysWebxia. A Text Accessibility Model for
People with Dyslexia. PhD thesis, Universitat Pompeu
Fabra, 2014.
[4] L. Rello, R. Baeza-Yates, and J. Llisterri. DysList: An
annotated resource of dyslexic errors. In Proc.
LREC’14, pages 1289–1296, Reykjavik, Iceland, 2014.
[5] L. Rello, C. Bayarri, and A. Gorriz. What is wrong
with this word? Dyseggxia: a game for children with
dyslexia (demo). In Proc. ASSETS ’12, pages 219–220,
Boulder, USA, October 2012. ACM Press.
[6] L. Rello, C. Bayarri, Y. Otal, and P. Pielot. A
computer-based method to improve the spelling of
children with dyslexia using errors. In Proc.
ASSETS’14, Rochester, USA, October 2014.
[7] H. Remschmidt and S. V. Aster. Kinder- und
Jugendpsychiatrie. Eine praktische Einf¨
uhrung: 177
Tabellen. Thieme, Stuttgart [u.a.], 2005.
[8] G. Schulte-K¨
orne, W. Deimel, and Jungermann, M.
and Remschmidt, H. Nachuntersuchung einer
stichprobe von lese-rechtschreibgest¨
orten kindern im
erwachsenenalter. Zeitschrift f¨
ur Kinder und
Jugendpsychiatrie und Psychotherapie, 31(4):267–276,
2003.
[9] P. H. K. Seymour, M. Aro, and J. M. Erskine.
Foundation literacy acquisition in European
orthographies. British Journal of Psychology,
94(2):143–174, 2003.
[10] C. Sterling, M. Farmer, B. Riddick, S. Morgan, and
C. Matthews. Adult dyslexic writing. Dyslexia,
4(1):1–15, 1998.
4https://itunes.apple.com/us/app/dyseggxia/
id534986729?l=ca&ls=1&mt=8