F
RISIAN READING COMPREHENSION
WITHIN PRIMARY EDUCATION
A
BOUT THE BENEFITS OF TRILINGUAL EDUCATION AND SPEAKING
F
RISIAN
ON
F
RISIAN READING COMPREHENSION SKILLS
N.A. Varkevisser
Student number: 2775328 Master Thesis Multilingualism Department of Applied Linguistics &
Department of Frisian Language and Culture Faculty of Arts, University of Groningen
Supervisor: Dr H.I. Hacquebord
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism
T
ABLE OF CONTENTS
Table of contents ... 2 List of Tables ... 3 List of Figures... 3 Summary ... 4 Samenvatting... 4 Gearfetting ... 5 1. Introduction ... 6 1.1 Current study... 6 1.1.1 Research questions ... 7 1.1.2 Expectations ... 8 1.1.3 Points of consideration ... 9 1.2 Reading guidelines ... 10 2. Background ... 122.1 Frisian in primary education ... 12
2.1.1 Historical overview ... 12
2.1.2 Trilingual education... 13
2.1.3 Initiatives for teachers ... 15
2.2 Frisian within the home situation ... 16
2.2.1 Historical overview ... 16
2.2.2 Recent initiatives ... 16
2.3 Reading comprehension ... 17
2.3.1 Defining reading comprehension ... 17
2.3.2 Differences in reading comprehension ... 18
2.3.3 Frisian reading comprehension ... 19
3. Method ... 21 3.1 Participants ... 21 3.2 Materials ... 23 3.3 Procedure ... 24 3.4 Plan of analysis ... 25 4. Results ... 26
4.1 Relationship between Frisian and school types ... 26
4.2 Differences on Frisian between language backgrounds ... 27
5. Conclusion and Discussion ... 29
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism
5.1.1 Points of consideration ... 29
5.1.2 Relationship between Frisian and school types ... 29
5.1.3 Differences on Frisian between language background... 30
5.2 Remarks ... 31
5.3 Recommendations ... 31
References ... 34
Appendix I: Frisian core objectives throughout the years ... 37
Appendix II: Requirements for trilingual schools ... 39
Appendix III: Questionnaire ... 40
Appendix IV: Norm scales ... 41
Appendix V: Screenshot ... 42
Appendix VI: Variables ... 43
L
IST OF
T
ABLES
Table 1: The minimum required amount of time spent on Frisian as subject and as
medium of instruction for each group in each trilingual school type in minutes per week 14
Table 2: The target levels (A1, A2, B1) for the language domain ‘reading’ divided by
group in each trilingual school type 15
Table 3: General background information of the complete group of participants and
per separate primary school group 21
Table 4: General background information of the complete sample of this master thesis
and per separate primary school group 23
Table 5: Outcomes of the Mann-Whitney U-tests for comparing groups with different
language backgrounds 28
L
IST OF
F
IGURES
Figure 1: Amount of primary schools participating in the Network of Trilingual Schools
from 2006 onwards. 13
Figure 2: Schematic of Frisian reading abilities of the inhabitants of the province
Fryslân in 2007, 2011 and 2014 in percentages. 17
Figure 3: Mean, median and mode scores on the Frisian reading comprehension test
divided by regular schools and certified schools. The lower the value in the graph, the
higher the score on the Frisian reading comprehension test. 26
Figure 4: Mean, median and mode scores on the Frisian reading comprehension test
divided by the five school types. The lower the value in the graph, the higher the score
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism
S
UMMARY
For a good participation in the society, it is important that children learn to read and understand what is written to gain the information they need. In Fryslân, information can be provided in both Dutch and Frisian. Primary schools are therefore obligated to teach Frisian to their students. Six core objectives have been formulated for Frisian, which indicate the skills that children should have when they are leaving primary school. In this study, Frisian reading comprehension is investigated among students from group 6, 7 and 8. It is examined if a relationship can be found between the scores on a Frisian reading comprehension test and the (trilingual) school type the students attend. Furthermore, it is examined if differences can be found between students who have different language backgrounds. In total 352 students of 22 different rural schools participated. A very small correlation was found between the scores and the (trilingual) school type. This slightly indicates that students from trilingual schools obtain higher scores on Frisian reading comprehension compared to students from regular schools, but must be considered negligible and future research is needed to strengthen this finding. In addition, statistical significant differences were found between Frisian students and Dutch and multilingual students. This indicates that Frisian students obtain higher scores on Frisian reading comprehension compared to Dutch and multilingual students, without considering the school type these children are attending. The conclusion that can be drawn from these results is that speaking Frisian at home with parents is beneficial for the Frisian reading comprehension of these students. It is therefore important that the initiatives that are taken to stimulate parents to use Frisian at home (e.g. the language bag) are continued.
KEY WORDS: Frisian; Language background; Multilingualism; Primary education; Reading
comprehension; Trilingual schools
S
AMENVATTING
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism spreken met beide ouders een positief effect heeft op begrijpend lezen in het Fries van deze leerlingen. Het is dan ook belangrijk dat de initiatieven die genomen worden om ouders te stimuleren om Fries te spreken (vb. taaltas) doorgezet worden.
KERNBEGRIPPEN: Begrijpend lezen; Drietalig onderwijs; Fries; Leesbegrip; Meertaligheid; Primair
onderwijs; Taalachtergrond
G
EARFETTING
Om goed partisipearje te kinnen yn de maatskippij is it wichtich dat bern lêze leare, mar ek goed begripe wat se lêzen hawwe. Sa kinne se de ynformaasje dy’t se nedich hawwe efterhelje. Yn Fryslân kin de ynformaasje sawol yn it Nederlânsk as yn it Frysk oanbean wurde. Basisskoallen binne dan ek ferplichte om omtinken te jaan oan it Frysk. Der binne seis kearndoelen foar it Frysk formulearre dy’t oanjouwe wat bern kinne moatte oan de ein fan de basisskoalle. Yn dit ûndersyk is sjoen nei it Frysk lêsbegryp fan learlingen út groep 6, 7 en 8. Earst is ûndersocht oft der in ferbân is tusken de skoares dy’t learlingen op in Fryske lêsbegryptoets helje en it type (trijetalige) skoalle dêr’t se hinne geane. Dêrnei is sjoen oft der ferskillen binne tusken learlingen mei ferskate taaleftergrûnen. Yn totaal hawwe 352 learlingen fan 22 ferskillende plattelânskoallen meidien oan it ûndersyk. Der waard in tige lytse korrelaasje fûn tusken de skoares op de Fryske lêsbegryptoets en it type (trijetalige) skoalle. Dat betsjut dat learlingen fan trijetalige skoallen hegere skoares helje op de Fryske lêsbegryptoets yn ferliking ta learlingen fan reguliere skoallen. Dit resultaat is eins net it neamen wurdich, mear ûndersyk is nedich om dit resultaat better te ûnderbouwe. Fierder waard der in statistysk signifikant ferskil fûn tusken Fryske learlingen en Nederlânske en meartalige learlingen. Dat betsjut dat Fryske learlingen hegere skoares helje op de Fryske lêsbegryptoets yn ferliking ta Nederlânsktalige en meartalige learlingen. De konklúzje dy’t hjirút lutsen wurde kin, is dat thús Frysk prate mei beide âlden in posityf effekt hat op it begripend lêzen Frysk fan dizze learlingen. It is dan ek wichtich dat de inisjativen om âlders te stimulearjen Frysk te praten (bygelyks it taaltaske) trochset wurde.
KEARNBEGRIPEN: Begripend lêzen; Frysk; Lêsbegryp; Meartalichheid; Primêr ûnderwiis;
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism
1.
I
NTRODUCTION
Eventually it is important that children learn to read and understand what they have read, since participating in the society would otherwise become quite difficult (Robbe, 2007). In the province Fryslân, in the northern part of The Netherlands, children not only have to learn to read in the majority language, Dutch, but children are also expected to learn to read in the minority language, Frisian (see Appendix I). It is, however, problematic to test whether the children are proficient in Frisian, since there are only a few tests available. This was also one of the problems the inspectorate bumped into when examining how Frisian is taught in schools (Inspectie van het Onderwijs, 2010).
Even though there are not that many possibilities to check whether children master Frisian, teaching Frisian in the province is seen as very important. Taalsintrum Frysk (2005) gives four motives why Frisian should be taught in the province Fryslân. First, there is the pedagogical-didactic motive, which is especially important for children who speak Frisian at home. For these children, Frisian is the most personal and familiar language. If only Dutch would be used in education, they might get an unsafe feeling that would not be beneficial for their (language) development. Although this argument cannot be put forward for non-Frisian children, teaching in Frisian might also be seen as beneficial for them. Namely, these children will get the chance to become multilingual at an early age, which is positive for their language development. The second motive is a cultural motive, since Frisian, as spoken and written language, can be seen as cultural heritage. Teaching Frisian is important for transmitting this cultural heritage. The third motive is a social motive and focusses on the fact that the province Fryslân is bilingual. Therefore, it would be desirable that every citizen would at least master Frisian in a passive way. The final motive is based on language politics. Dutch is the majority language in The Netherlands and is (more) supported by the government. To keep Frisian alive, it is important that it gets/has an equal place in the society. Education can give this a positive impulse (Taalsintrum Frysk, 2005).
Several initiatives are taken to strengthen the position of Frisian within society. One of these initiatives in education is the formation of trilingual primary schools. Within these schools, Frisian is a language of instruction for subjects just as Dutch and English are (Cedin, 2013). Other initiatives within education are, for example, bilingual playgroups and crèches (Dutch/Frisian) and trilingual secondary schools (Provinsje Fryslân, 2012). These initiatives make it possible to create a continuous learning path for children in which the Frisian language is included. To provide the best possible teachers for these multilingual schools, there is a special multilingual teacher-training program (Wolf, 2013). In addition, initiatives outside education are set up to provide a better position for Frisian, such as the magazine Heit en Mem, which is a free bilingual magazine (Dutch/Frisian) for parents with children aged 0-12. The magazine provides information about multilingualism and about how to raise children with more than one language (Bruinsma, 2015). Furthermore, new parents are provided with a language bag, Taaltaske, from the municipality government. This bag is filled with information about multilingualism and about how to raise children with more than one language (Provinsje Fryslân, 2014). For adults, language courses are available on all sorts of levels and language domains to get a better understanding of the Frisian language (Afûk, 2015).
1.1
C
URRENT STUDY
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism Using the Frisian language within education is seen as important. This derives, inter alia, from the special core objectives that are set up for Frisian. One of these core objectives focusses on the written form: children in the province Fryslân are expected to read in Frisian and then, of course, understand what has been written. One of the difficulties with investigating this is that there are not that many tests available. During the master internship, the author of this thesis was involved in the development of new reading comprehension tests for Frisian for the higher groups of primary education. This provided a well-based set up for further investigation of the results of these tests in a master thesis. In this master thesis, the results will be combined with the trilingual school system and the language background of the children. Are all these initiatives paying off considering Frisian reading comprehension?
1.1.1
RESEARCH QUESTIONS
The first initiative that will be investigated in combination with Frisian reading comprehension is the trilingual school system. This school system has been set up to strengthen the position of Frisian next to Dutch. Within these schools, Frisian is not only taught as subject but also used as language of instruction. In the school year 2013/2014, Bleker and De Jager (2015) investigated whether students of group 8 of trilingual primary schools attained higher achievements on, inter alia, Frisian reading comprehension compared to students of group 8 of regular primary schools1. It turned out that students on trilingual schools indeed have higher scores on Frisian reading comprehension, tested with the CITO-test Begripend lêzen Frysk, compared to students of regular schools (Bleker & De Jager, 2015).
In the study of Bleker and De Jager (2015) only students from group 8 were tested. In this master thesis, the results from students from group 6, 7 and 8 will be investigated. The first research question formulated, based on this initiative, is:
1. Do students from group 6, 7 and 8 of trilingual primary schools get higher scores on Frisian reading comprehension compared to students from group 6, 7 and 8 from regular primary schools?
The second initiative that will be investigated in combination with Frisian reading comprehension is the influence of the language parents speak to their children. Inter alia, the magazine Heit en Mem and the language bag, Taaltaske, stimulate parents to raise their children with more than one language. Quite recently, Provinsje Fryslân (2014) found that 88% of the parents also see raising children with more than one language as a positive outcome for the development of the children. However, 75% of the parents in this research also mentioned that their children changed their spoken language since they attained school. Children who were raised in Frisian or bilingual Dutch/Frisian started to use more Dutch at home since they went to school (Provinsje Fryslân, 2014). Unfortunately for this master thesis, the target group in the research of Provinsje Fryslân (2014) existed only of parents with children in group 1 and 2 of primary education. It is difficult to say if this trend in using more Dutch will also continue when children grow older and what the influence will be on their Frisian reading comprehension. Van Ruijven and Ytsma (2008) investigated the profits of the trilingual educational model in the province. They found that there is no difference between Frisian speaking students and Dutch speaking students in either group 6, 7 or 8 on Frisian reading comprehension on either regular primary schools or trilingual schools (Van Ruijven & Ytsma, 2008).
In the study of Van Ruijven and Ytsma (2008), the trilingual school system had just taken off; using three languages within education was quite new. One aspect they did not clarify in their
1
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism research was the definition of a Frisian and Dutch child; they only mentioned the word mother
tongue. The ‘problem’ with Frisian speaking people is that they should all be considered
bilinguals, since they are expected to speak at least Dutch next to Frisian. Since a distinction is made between Frisian and Dutch students in this master thesis, it is seen as important to clarify when someone is considered Frisian, Dutch or bi- or multilingual. Therefore, the following definition has been formulated:
A student is considered Frisian when (s)he speaks Frisian with both parents. A student is considered Dutch when (s)he speaks Dutch with both parents. A student is considered multilingual when (s)he speaks a different language with each parent.
The definition given above arises from the definition given by Grosjean (2010) about bilinguals: “Bilinguals are those who use two or more languages (or dialects) in their everyday life” (p. 4). This definition is narrowed for this master thesis, only including the language contact with parents. Furthermore, it is referred to as multilingual as this ‘multi’ strengthens the more than
one language term better. Logically, this means that monolingual Frisian or Dutch speakers must
be those who use only one language (or dialect) in their everyday life with their parents. The definition was narrowed on purpose, since the students from group 6, 7 and 8 (± 9-12 year-olds) will be expected to indicate whether they are monolingual or multilingual themselves. If they have to include all the contacts they have on one day, it might lead to confusion. Combining all these aspects, the second research question is formulated as:
2. Do Frisian students from group 6, 7 and 8 get higher scores on Frisian reading comprehension compared to Dutch and multilingual students from group 6, 7 and 8, irrespectively of a trilingual or regular school background?
1.1.2 EXPECTATIONS
Since Bleker and De Jager (2015) have performed a comparable research, it is expected that the results on the first research question will be in accordance with their findings. The expectation for the first research question is therefore that students of trilingual schools will achieve higher scores compared to students from regular schools. Nonetheless, Bleker and De Jager (2015) only tested students from group 8 and used the CITO-test Begripend lêzen Frysk. Whereas for this master thesis group 6, 7 and 8 will be tested with newly developed Frisian reading comprehension tests.
Despite that Van Ruijven and Ytsma (2008) found no differences between Frisian and Dutch students on Frisian reading comprehension, neither on regular nor on trilingual schools, it can be different in this case. Van Ruijven and Ytsma (2008) gathered data in the school years 2001 up until 2006, which means that ten years have passed. Moreover, not only has the trilingual school system improved, but also more initiatives have been taken to raise children with more than one language with no shame in using Frisian as one of the languages. Therefore, the expectation for this master thesis is that Frisian students will get higher achievements on Frisian reading comprehension compared to students who only speak Dutch with their parents or students who speak a different language with each parent. Logically, this is expected because the immersion in Frisian has been longer and more intense for Frisian speaking children. There is another reason why the expectation of this research is not linked to the findings of Van Ruijven and Ytsma (2008). It is not clear what Van Ruijven and Ytsma (2008) defined as Frisian and as
Dutch whereas for this research a clear definition is formulated. It is not known if these
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism research, while the chances are that they were participating in the research. It is however not clear in which group they were participating. To avoid confusion it was therefore important to give a clear definition of the language background of the participants in this master thesis.
1.1.3 POINTS OF CONSIDERATION
There are some points of consideration about the concept of reading, before the results on the Frisian reading comprehension tests can be interpreted. Having a good understanding of what is read; different skills must be mastered before that is even possible. Beernink, Van Koeven, Litjens, De Groot and Vreman (2006) explain that before one can understand what (s)he is reading one needs to be able to read more or less fluently. When one reads fluent, there is time and space to pay attention on what is read, combining word groups, sentences and paragraphs. Reading strategies can help to get a better understanding of what is read and good readers will apply the learned strategies automatically when needed (Beernink et al., 2006). In the case of reading, most schools in the province Fryslân, if not all, start with Dutch reading (mostly in group 3). For the children who speak Dutch at home with both parents, this would be learning to read in their first language. However, for children who speak Frisian at home with both parents and learned Dutch outside the home situation, this would be learning to read in their second language. According to Kurvers and Mooren (2005), it would be best for children to learn reading in the language that is most familiar to them, in other words: their mother tongue. The teacher plays in this case a crucial role, as some children will read words that they might not have used actively before if Dutch is not their mother tongue.
As it is one of the core objectives, schools will introduce reading in Frisian at a certain moment. This happens after children have learned to read in Dutch. For children who speak Dutch at home this will be learning to read in a second language after they have gained knowledge of how to read in their mother tongue. On the other hand, this would be learning to read in their mother tongue for children who speak Frisian at home, who have knowledge of how to read in a second language. Oosterloo (2004) explains that it is expected from students that they use the knowledge they have learned in reading Dutch on reading in Frisian. Students who have difficulties applying the right reading strategy at the right moment in Dutch, might also fail to do this in Frisian. In addition, it might even be questioned if students are able to read the same text types in Frisian and apply the (Dutch) reading strategies on these Frisian texts. Students learn to read Frisian later and it might therefore be assumed that their knowledge of written Frisian is smaller compared to their knowledge of written Dutch.
In addition, reading disorders also have influence on one’s reading abilities. Children with dyslexia, for example, have difficulties with fluent and accurate reading and spell on word level (Ghesquière & Ruijssenaars, 2008). These children might need extra time because reading and spelling goes slower, but they might also make more mistakes due to their dyslexia. The more complex the language task becomes, the more (time) pressure is felt and problems might get worse (Beernink et al., 2006). Reading in Frisian might be seen as more complex. For Dutch children it will be learning to read in a second language, after they have learned to read in their mother tongue, which already can be complex if you are dealing with dyslexia. Although it might be the mother tongue language for Frisian students, they too have only learned to read Dutch in school and might have (extra) difficulty with that due to their dyslexia. If any of these children already feel pressure and have problems during Dutch reading, they might have these feelings even more during Frisian reading.
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism perform better at mathematics. For example, Logan and Johnston (2009) found for ten-year-olds that “girls had better reading ability, read more frequently and had a more positive attitude to
reading and school compared with boys” (p. 207). This assumption and quote might also apply on
the students tested in this master thesis.
1.2
R
EADING GUIDELINES
In order to give a structured answer on the two research questions, this master thesis is divided into four chapters. Below a brief guideline is given of what will be highlighted in these chapters.
BACKGROUND
Using Frisian within official and public domains was not always possible, especially in education it was not allowed to speak Frisian. Nowadays, there is a growth in initiatives to strengthen the position of Frisian within these domains. One of these initiatives is the trilingual school system, where Dutch, English and Frisian are used as languages of instruction. As this has not always been the case, a historical overview of how Frisian gained a position in education is described first, followed by an explanation of the trilingual school system and the initiatives that are taken for (soon-to-be) teachers.
Since Frisian was not always part of the educational system, parents transferred Frisian at home. An overview of how Frisian is used in the home situation is given. Furthermore, the results of the initiatives that are taken for the home situation, such as the language bag, are briefly described.
The last topic described in this chapter deals with the overlapping topic of both research questions: reading comprehension. First, a general description of reading comprehension is given. Then the points of consideration that were mentioned, dyslexia and gender differences, are further investigated. The chapter ends with an explanation of how Frisian reading comprehension can be studied.
METHOD
To give an answer to the research questions, data has been gathered of 352 students from 22 different schools. In this chapter, the background information is given of these students and the schools they are attending. In addition, the points of consideration that were mentioned, dyslexia and gender differences, are investigated to check if differences for these groups are found.
Furthermore, a description is given of the materials that have been used in this study. One of the materials is a questionnaire that has been developed by the author of this thesis. Moreover, the Frisian reading comprehension tests that have been used are developed by the lectureship Frisian and Multilingualism of the universities of applied sciences in Leeuwarden. The tests are briefly discussed in this chapter.
Additionally, an overview of the statistical methods that will be used to answer the research questions is given.
RESULTS
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism The second research question can be characterized as a question searching for a difference between different variables. Since one variable can be considered nominal and the other ordinal, a Kruskal-Wallis Test and a Mann-Whitney U-test were used.
CONCLUSION AND DISCUSSION
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism
2.
B
ACKGROUND
Using Frisian within official and public domains was not always possible, especially in education it was not allowed to speak Frisian. Nowadays, there is a growth in initiatives to strengthen the position of Frisian within these domains. One of these initiatives is the trilingual school system, where Dutch, English and Frisian are used as languages of instruction. As this has not always been the case, a historical overview of how Frisian gained a position in education is described first, followed by an explanation of the trilingual school system and the initiatives that are taken for (soon-to-be) teachers.
Since Frisian was not always part of the educational system, parents transferred Frisian at home. An overview of how Frisian is used in the home situation is given. Furthermore, the results of the initiatives that are taken for the home situation, such as the language bag, are briefly described.
In the last paragraph of this chapter a general description of reading comprehension is given. In addition, the points of consideration that were mentioned in the introduction, dyslexia and gender differences in reading (see chapter 1.1.3 Points of consideration), are further investigated. The paragraph ends with an explanation of how Frisian reading comprehension can be studied.
2.1
F
RISIAN IN PRIMARY EDUCATION
2.1.1
HISTORICAL OVERVIEW
Nowadays, it is quite normal to hear the minority language Frisian as language of instruction in primary schools in the province Fryslân in The Netherlands. Despite that, this has not always been the case. Some changes in the last decades have led to an increase of using Frisian within education. According to Ytsma, Riemersma and De Jong (2010) the use of Frisian ‘in education’ dates back to 1907. In this year, the provincial government contributed with funding for Frisian lessons after the regular school hours, making it an extra-curricular subject. It became possible to teach Frisian in the higher grades during the Dutch lessons when the Lager Onderwijs Wet (Primary Education Act) was altered in 1937 (Ytsma, Riemersma, & De Jong, 2010).
In 1950, an experiment with bilingual primary education started, which resulted into a legal basis for bilingual schools in 1955. From then on, Frisian was allowed as medium of instruction in the lower grades and teaching Frisian as optional subject became possible throughout primary school. The Primary Education Act was changed again in 1974; this approved the use of Frisian as medium of instruction throughout primary school instead of only in the lower grades. From 1980 onwards, Frisian became an obligatory subject in primary education (Ytsma, Riemersma, & De Jong, 2010). In 1993, core objectives were established for all school subjects, including specific core objectives for Frisian (Staatsblad, 1993). The complete set of core objectives for all primary school subjects were revised in 1998 (Adelmund, 1998), nonetheless, the Frisian core objectives stayed the same as those of 1993. Until 2006, the Frisian core objectives mentioned by Staatsblad (1993) and Adelmund (1998) equalled those of Dutch, but in 2006, the Frisian core objectives were revised and limited to the way they are now (Inspectie van het Onderwijs, 2006). In Appendix I, an overview of the core objectives for Frisian is given. It also shows the limitation from 12 core objectives in 1993/1998 to six core objectives nowadays.
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism certificate, it is assumed that the school uses not only Dutch, but also Frisian and English as medium of instruction on a regular basis (see chapter 2.1.2 Trilingual education). However, this does not mean that every school in the province Fryslân is obligated to become a certificated trilingual school. Nevertheless, as Riemersma and De Vries (2011) mention, all schools in Fryslân are at least obligated to teach Frisian (since 1980) and English (since 1986) as a subject. Sometimes, schools can ask for an exemption for teaching Frisian as subject. The exemption is granted if, for example, less than five percent of the student population has a Frisian background (Rijksoverheid, 2015). These schools are, for example, located on the Wadden Islands, the municipality Het Bildt and the municipalities Ooststellingwerf and Weststellingwerf.
While schools are obligated to teach Frisian, the amount of time spent on the subject varies because schools are allowed to schedule Frisian themselves. The inspectorate showed that in regular schools the groups 1 and 2 receive an average amount of 30 minutes of Frisian per week, and the groups 3 to 8 receive between 30 and 45 minutes of Frisian per week. In a complete school year, this would be approximately 20 hours for group 1/2 and 20-30 hours for group 3 to 8 (Inspectie van het Onderwijs, 2006). In 2010, a new report of the inspectorate was published,
Tusken winsk en wurklikheid, which revealed that the amount of time spent on Frisian as subject
seems to rise, especially in the lower grades of primary school (Inspectie van het Onderwijs, 2010). However, the average time spent is calculated from the sum of time spent on regular schools and trilingual schools. Unfortunately, it is therefore not possible to use these averages to give an indication of how much time is spent on Frisian as subject in regular schools nowadays.
2.1.2
TRILINGUAL EDUCATION
As mentioned above, many schools in Fryslân use Dutch, Frisian and English as medium of instruction nowadays. This is one of the requirements for obtaining a trilingual school certificate provided by Cedin/Taalsintrum Frysk (Cedin, 2011). According to Riemersma and De Vries (2011), the trilingual school project started in 1997 to improve the quality of education in particularly Frisian and English, as it turned out that the core objectives set for these languages were not reached. In 1997 five primary schools in the province Fryslân formed an experimental group where the trilingual school project was started. In 1998 two other schools followed. These seven schools successfully completed the project in 2006 by obtaining a trilingual school certificate and they joined the Network of Trilingual Schools (Riemersma & De Vries, 2011). In Figure 1, the growth2 of the Network of Trilingual Schools can be found. Each school within this Network is applying for a trilingual school certificate or already has a certificate.
2
Numbers provided by Riemersma and De Vries (2011; p. 46) and Taalsintrum Frysk/Cedin (2007; pp. 10-11) – 2006/2007; Wikipedy: De frije ensyklopedy (2011) – 2008; Riemersma and De Vries (2011; p. 49) – 2009; Goodijk (2010; p. 27) and Inspectie van het Onderwijs (2010; p. 16) – 2010; Benedictus-van den Berg (2012; p. 183) – 2011; Bleker and De Jager (2015; p. 3) and Wolf (2013; p. 29) – 2013; Bleker and De Jager (2015; p. 3) – 2014; Goodijk (2015) – 2015.
0 10 20 30 40 50 60 70 80 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism All requirements to become a trilingual school can be found in Cedin (2011) and in Appendix II. However, not every trilingual school has to be the same (Cedin, 2013). There are five different possible types (Cedin, 2013), which are all part of the Network of Trilingual Schools:
1. Startfariant meartaligens 1 (Start variety multilingualism 1), 2. Startfariant meartaligens 2 (Start variety multilingualism 2), 3. 3TS3 Basis (3TS Basic)4,
4. 3TS Plus (3TS Plus), and 5. 3TS Ster (3TS Star)5.
Using Dutch, Frisian and English as medium of instruction is one of the requirements for a trilingual school (Cedin, 2011). Nevertheless, the five trilingual school types all differ, inter alia, in the required amount of Frisian taught in school. In Table 1, a schematic overview of these different types of trilingual schools and the amount of time that they need to spent on Frisian as a subject and as medium of instruction is given. The results of the amount of time spent on Frisian found by the inspectorate (Inspectie van het Onderwijs, 2006) are comparable to the amount of time spent in a Start variety 1 school. Apparently, this amount of time is seen as minimum by Cedin (2013) and it might perhaps lower the threshold to become part of the Network.
Table 1: The minimum required amount of time spent on Frisian as subject and as medium of instruction for each group in each trilingual school type in minutes per week (Cedin, 2013)
Start variety 1 Start variety 2 3TS Basic 3TS Plus 3TS Star Group 1/2 Subject 30 min. 30 min. 60 min. 60 min. 60 min.
Medium 105 min. 150 min. 270 min. 40% of complete class time Group 3/4 Subject 45 min. 45 min. 60 min. 60 min. 60 min.
Medium 45 min. 75 min. 135 min. 210 min.
Group 5/6 Subject 45 min. 45 min. 60 min. 60 min. 60 min.
Medium 45 min. 75 min. 135 min. 210 min.
Group 7/8 Subject 45 min. 45 min. 60 min. 60 min. 60 min.
Medium 60 min. 75 min. 135 min. 210 min.
The pattern in Table 1 shows that the ‘higher in hierarchy’ of the trilingual school types, the more time is spent on Frisian, especially as medium of instruction. However, the descriptions given by Cedin (2013) are not clear on whether the language as medium of instruction only requires the spoken form or whether the written form has to be included as well. On the other hand, Cedin (2013) distinguishes four language domains (attitude & understanding; reading; speaking; writing) and the targets for these domains increase when one goes a step higher in the trilingual school type. In Table 2, the target levels for the domain ‘reading’ in Frisian and English per group in each trilingual school type are given. The pattern in this table shows that an increase in the proficiency level of reading in Frisian and English is expected when a school is ‘higher in hierarchy’ in the trilingual school system.
3
3TS means TrijeTalige Skoalle (Trilingual School).
4
When all requirements given by Cedin (2011) of 3TS Basic are achieved, the trilingual school certificate is provided.
5
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism Table 2: The target levels (A1, A2, B16) for the language domain ‘reading’ divided by group in each trilingual
school type (Cedin, 2013)
Start variety 1 Start variety 2 3TS Basic 3TS Plus 3TS Star
Group 1/2 A1 50% 60% 80% 60% - A2 - - 20% 40% 60% Group 3/4 A1 70% 80% 60% 50% - A2 - 20% 40% 50% 70% Group 5/6 A1 80% 60% 50% 50% - A2 10% 40% 50% 50% 80% B1 - - - - 20% Group 7/8 A1 70% 50% 30% - - A2 30% 50% 70% 80% 60% B1 - - - 20% 40%
2.1.3 INITIATIVES FOR TEACHERS
The trilingual school system (see chapter 2.1.2 Trilingual education) is of course a wonderful stimulation for use of the Frisian language within education. To provide the best possible teachers for these multilingual schools, a special multilingual teacher-training program is created. Within this program, soon-to-be teachers not only increase their own mastery of the different languages in the different language domains, but they also visit trilingual schools in other areas, such as Wales (Wolf, 2013). By looking at other possibilities of trilingual education, the trilingual education within Fryslân can be improved since the soon-to-be teachers will know the advantages and disadvantages of trilingual education in several situations. Furthermore, the multilingual teacher-training program also covers a minor, where the soon-to-be teachers learn all about multilingualism and the didactics behind it (Wolf, 2013).
This new multilingual teacher-training program is not (yet) obligatory for all soon-to-be teachers. The regular teacher-training programmes on the universities of applied sciences in the capital city of the province Fryslân, Leeuwarden, also pay attention to Frisian in education. Soon-to-be teachers improve their own proficiency in Frisian in all language domains (Stenden Hogeschool, 2011) and learn about didactics and multilingualism (Flapper & Tigchelaar, 2013). For teachers who have already graduated and are teaching at schools, there is the special refresher course for Frisian, paid by the provincial government. Here, teachers can improve their proficiency in Frisian in all language domains (Flapper, Wolf and Ybema, 2014) and learn all about multilingualism within education (Flapper & Ybema, 2010). Recently, the framework of reference for the Frisian language, referinsjeramt Frysk (rrF) written by Meestringa and Oosterloo (2015), has been published. The rrF can be compared to the Common European Framework of Reference for languages (CEFR) from the Council of Europe (2001). It gives tools to establish learning goals for Frisian, but also shows what someone already can-do in the language (Meestinga & Oosterloo, 2015). The rrF can help teachers and students with setting up learning goals for the students and show what they are already capable of doing with the language. Furthermore, it can help to shape the education that teachers provide because it can give insights in the aspects that need more time and further explanation.
6
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism
2.2
F
RISIAN WITHIN THE HOME SITUATION
2.2.1
HISTORICAL OVERVIEW
Using Frisian in official, public domains, such as in politics, law and education, was not allowed until the beginning of the twentieth century (Jonkman & Versloot, 2013). This means that Frisian was used in the home situation and more informal situations only, such as with friends. According to Jonkman and Versloot (2013) this changes scantily throughout the twentieth century with activities of the Fryske Beweging (Frisian Movement) in churches (1916), municipality councils (in the twenties), language education (see chapter 2.1 Frisian in primary education) and with the establishment of the Afûk7 (1928) and Fryske Akademy8 (Frisian Academy, 1938). As a result of Kneppelfreed (November 16th, 1951) it is allowed to use Frisian in official domains since 1956 and the use of Frisian as communication language is still increasing within institutions. Nonetheless, when someone lives in the province Fryslân they are not obligated to understand Frisian; it can only be seen as a favour towards the citizens who speak the language (Jonkman & Versloot, 2013).
2.2.2
RECENT INITIATIVES
Looking at the use of Frisian in more recent years, the different language atlases provided by Provinsje Fryslân (2007, 2011) give an indication. In the language atlases, it is found that a very large percentage of the respondents are able to understand spoken Frisian: about 84.2% in 2007 and 84.6% in 2011. Since reading comprehension is the main topic of concern in this master thesis, the Frisian reading abilities of the inhabitants of the province Fryslân in the years 2007 and 2011 are schematically given in Figure 2. The target group in the language atlases of the Provinsje Fryslân (2007, 2011) was adults of 18 years and older. Nevertheless, if this group is not able to read Frisian, their children are unlikely to be taught reading in Frisian by them. Therefore, it is rather positive to see the growth in Figure 2 over the years.
Quite recently, Partoer (2014) conducted a research among a Frisian citizen panel, Fries
burgerpanel: Fryslân inzicht, and found an increase in case of understanding Frisiancompared to
the results of Provinsje Fryslân (2007, 2011). Partoer (2014) found that about 92% of the citizens are able to understand Frisian as spoken language. According to Partoer (2014), there has been an increase in the Frisian reading abilities as well (see Figure 2). Again, the target group is adults of 18 years and older, but the same argument holds: if this group does not read the language, they will not be able to teach it to younger people.
7
These days, the mission of the Afûk is to bring Frisian close to all the people of the province, regardless of age, by means of language promotion and transmission. The Afûk is primarily a bookstore, and publishes and sells books mainly in Frisian. Furthermore, the Afûk provides Frisian courses on all sorts of levels and language domains, invests in information sessions and promotes Frisian through different projects, such as Tomke, Heit&Mem and Praat-mar-Frysk (see: http://www.afuk.frl).
8
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism Another positive finding for usage of Frisian in the home situation is the result of a research published by Provinsje Fryslân (2014). It was found that 88% of the questioned parents see raising children with more than one language as a positive development for their children. These results can be attributed to the initiatives that are taken to show parents the positive sides of multilingual upbringing. One of these initiatives is the language bag, Taaltaske, provided by the municipality government. This bag is filled with information about multilingualism and how to raise children with more than one language (Provinsje Fryslân, 2014). Other initiatives that are focused on parents are the spread of free magazines, such as Heit en Mem, that give information about multilingualism (Bruinsma, 2015). All these initiatives contribute to a positive attitude towards multilingual upbringing and not being ashamed of using Frisian, as minority language, as one of these languages. Riemersma and De Vries (2011) also concluded that the attitude of parents towards Frisian is quite positive, saying that ‘85% of the parents would find it a pity if the
Frisian language were to disappear (p. 54). A couple of years earlier, however, this attitude was
completely different. In 2004 the newspaper Dagblad van het Noorden wrote an article with the headline: ‘Het Fries zit zwaar in de verdrukking: Ouders denken dat onderwijs in tweede taal
ontwikkeling tegenwerkt – Onverschilligheid neemt toe.’ 9 In the article it is described that parents prefer to raise their children with only the Dutch language, since they believe that using Frisian will work against a good proficiency in Dutch and therefore stand in the way of a good development. The argument given in the article is the ignorance of parents about mulitlingual upbringing. Parents do not know what the benefits are for their children if they are raised with more than one language in the home situation and therefore only use Dutch (Stienstra, 2004). The above mentioned results show that this ignorance towards multilingual upbringing has, inter alia due to all the initiatives, been decreased.
2.3
R
EADING COMPREHENSION
2.3.1 DEFINING READING COMPREHENSION
When children are able to read more and more fluently, it is possible to pay attention to understanding what has been written. Reading comprehension can be defined as:
“Reading is the act of constructing meaning from text. We use skills, strategies and prior knowledge, all of which are developmental in nature, to understand what we read. The act of reading is supported by reader motivation and positive reader affect. We read to help us achieve our goals, within and outside the school”
(Afflerbach, 2012, p. 14).
9
Frisian is heavily oppressed: Parents think that education in a second language works against the development – Indifference increases. 25% 29% 46% 24% 27% 49% 14% 23% 63% 0% 10% 20% 30% 40% 50% 60% 70%
With difficutly/not Reasonable (Very) well
2007 2011 2014
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism Verhoeven and Perfetti (2008) explain that the first skill necessary for understanding a text is the ability to identify individual words and decoding what these words mean: “’known’ forms are
recognized more or less directly, rare or complex forms are often recognized via parsing or the segmentation of the word in its morphological constituents” (p. 295). The second skill is to
understand the sentences wherein the identification of words plays a crucial role: “Sentence
comprehension can at best be understood as an operation which uses both sentence structure and word meanings to formulate hypotheses about the meaning of the sentence” (p.295). The last skill
is the complete text comprehension, where the reader must combine the meanings of all the sentences (s)he read. However, not only the information given in the text is needed for full text comprehension, also prior knowledge must be activated (Verhoeven & Perfetti, 2008).
Furthermore, several strategies can help to get a better understanding of a text. Beernink and colleagues (2006) mention seven strategies:
1. Determination of the goal of the text; 2. Recognition of the text type;
3. Determination of the main idea of the text;
4. Determination of the main idea of each paragraph; 5. Determination of the meaning of difficult words; 6. Understand the reference of words and sentences; 7. Judge a text.
Readers are expected to be curious of what the information in the text has to offer them. A reader will search for answers in a text, especially when specific questions about a text are asked. In addition, Beernink and colleagues (2006) emphasize the importance of reading motivation because this can help to develop the reading strategies. According to Van Coillie (2007) reading motivation can be established by humour and thrilling in a text. Additionally, connection to the interests of the reader can provide a higher reading motivation.
2.3.2 DIFFERENCES IN READING COMPREHENSION
Teachers will always try to teach reading skills and strategies and increase the reading motivation of children, which will help them with reading and understanding a text. Nonetheless, sometimes other aspects influence reading comprehension. For example, differences in reading comprehension proficiency can occur when someone deals with a reading disorder. Different reading disorders can be distinguished whereof dyslexia is probably the most known. Ghesquière and Ruijssenaars (2008) explain that dyslexia can be characterized by having persistent problems in learning and the accurate and fluent applying of reading and spelling on word level. Students with dyslexia can be recognized by a large unexpected difference between their potential school success and their actual school performance. However, this difference varies enormously between individuals with dyslexia (Ghesquière & Ruijssenaars, 2008). Beernink and colleagues (2006) explain that one of the tools that might help children with dyslexia is to give them extra time to do the reading. When dyslectic children feel a lot of time pressure while reading, they will make more mistakes than necessary. Due to their dyslexia, children already make more mistakes compared to children without dyslexia. Therefore, if extra time can decrease this a little bit, it would make a big difference for these children. Van Coillie (2007) explains that children with dyslexia often guess what is written, which also leads to mistakes.
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism obligatory exam subjects and therefore exemptions for these subjects cannot be made. Frisian, however, is not an obligatory exam subject. A school can decide to give children an exemption for Frisian when they have dyslexia, especially for reading and writing. In many cases, this means that the language domains speaking and listening can still be followed (Steunpunt Dyslexie, 2014). It is not known how many schools give this exemption to their dyslectic students.
Another example of when differences in reading comprehension can occur depends on gender. Nowadays for example, people often say that men have better mathematic skills compared to women, and women have better language skills than men. Support for this statement is found in literature. Looking at the early development of children, Berk (2009) points out that girls go ahead of boys in language progress. Girls start to talk earlier and show faster vocabulary growth later on. Girls also achieve higher scores on reading abilities and have fewer need of remedial reading instructions compared to boys. Berk (2009) gives several explanations, such as “girls show a biological advantage in earlier development of the left
hemisphere of the cerebral cortex” (p. 551) and “mothers also talk considerably more to girls than to boys” (p. 551). Furthermore, Logan and Johnston (2009) found for ten-year-olds that “girls had better reading ability, read more frequently and had a more positive attitude to reading and school compared with boys” (p. 207). However, the significant differences in their research were
quite small.
The aspects mentioned above are very important when it comes to testing children on their reading comprehension skills. If, for example, significant differences are found between boys and girls, it might be better to split them up for further analyses. The same holds for finding significant differences between children with and without reading disorders. If this goes unnoticed, wrong conclusions can be drawn which might, for example, influence the didactics used in the classroom. In this master thesis, both gender and dyslexia will be borne in mind to make sure that these aspects do not influence the results.
2.3.3 FRISIAN READING COMPREHENSION
Frisian reading comprehension can also be seen in terms of the definition given by Afflerbach (2012). It is intended that readers construct the meaning of a text written in the Frisian language. Schools have the possibility to test the students on Frisian reading comprehension in different ways, for example by using the tests in the Frisian method StudioF. Using this method is one of the requirements for schools within the Network of Trilingual Schools (Cedin, 2013). Different language domains are covered in this method and can be tested by using the tests belonging to this method, Frisian reading comprehension is one of them starting from group 4 onwards. However, Oosterloo (2004), who wrote the blueprint for StudioF, advices to use a different method for reading and listening comprehension: namely the method Begripen harkjen
en lêzen of GCO-Fryslân (2000). The downturn of this method is, however, that there are no tests
available. Furthermore, results on how students scored on the reading comprehension test of StudioF are not published.
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism of regular schools.
Other ways to indicate how students in primary schools in the province Fryslân perform in Frisian reading comprehension are not available. No public information is accessible about the tests from StudioF and there are no other tests for each school year available. This is one of the reasons why the lectureship Frysk en Meartaligens yn opfieding en ûnderwiis (lectureship Frisian and Multilingualism in upbringing and education; lectureship FMOO) has started to develop new Frisian reading comprehension tests. These tests are not bound to a method and give insights in the Frisian reading comprehension for students from primary school to secondary school. These tests can therefore be called norm-referenced tests. According to Van den Brink and Mellenbergh (2003) this indicates that the results of the students are not compared to the educational objectives given in the method, but to each other. The norm that is used within these tests makes it possible to create an ordering of higher and lower scores and compares these with respect to each other (Van den Brink & Mellenbergh, 2003). The norms for the tests developed by the lectureship FMOO can be found in Appendix IV.
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism
3.
M
ETHOD
The two research questions in this study are as follows:
1. Do students from group 6, 7 and 8 of trilingual primary schools get higher scores on Frisian reading comprehension compared to students from group 6, 7 and 8 from regular primary schools?
2. Do Frisian students from group 6, 7 and 8 get higher scores on Frisian reading comprehension compared to Dutch and multilingual students from group 6, 7 and 8, irrespectively of a trilingual or regular school background?
To give an answer to the research questions, data has been gathered of 352 students from 22 different schools. In this chapter, the background information is given of these students and the schools they are attending. In addition, the points of consideration that were mentioned in the introduction, dyslexia and gender differences in reading (see chapter 1.1.3 Points of consideration), are investigated to check if differences for these groups are found.
Furthermore, a description is given of the materials that have been used in this study. One of the materials is a questionnaire that has been developed by the author of this thesis. Moreover, the Frisian reading comprehension tests that have been used are developed by the lectureship Frisian and Multilingualism of the universities of applied sciences in Leeuwarden. These tests are briefly discussed in this chapter.
Additionally, an overview of the statistical methods that will be used to answer the research questions is given.
3.1
P
ARTICIPANTS
In this study, 387 children of 22 different schools10 in the province Fryslân in The Netherlands participated. Unfortunately, no schools from the cities wanted to participate in the research, so all the participating schools are located in the rural area of the province. The children are all students of group 6, 7 or 8, the highest groups of primary education. In total, 132 students of group 6, 131 students of group 7 and 124 students of group 8 have participated. Among the schools, two regular schools, five schools with Start variety 1, one school with Start variety 2, 12 schools with a Basic trilingual school certificate and two schools with a Plus trilingual school certificate were involved. Among the participants, 198 boys and 168 girls have taken part in the tests. The age ranged between 9 and 13 years old, with an average of 10.60 years old (SD = .98). In Table 3, a detailed description of the general background of the participants is found of both the complete group and per separate group.
Table 3: General background information (school type, gender, age, language background) of the complete group of participants and per separate primary school group
Total (N = 384*) Group 6 (N = 131*) Group 7 (N = 129*) Group 8 (N = 124)
School type None 3TS 30 (7.8%) 25 (18.9%) 3 (2.3%) 2 (1.6%) Start var. 1 92 (23.8%) 10 (7.6%) 36 (27.5%) 46 (37.1%) Start var. 2 20 (5.2%) 0 (0.0%) 9 (6.9%) 11 (8.9%) 3TS Basic 215 (55.6%) 85 (64.4%) 72 (55.0%) 58 (46.8%) 3TS Plus 30 (7.8%) 12 (9.1%) 11 (8.4%) 7 (5.6%) 10
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism Gender Boy 198 (51.6%) 68 (51.9%) 76 (58.9%) 54 (43.5%) Girl 186 (48.4%) 63 (48.1%) 53 (41.1%) 70 (56.5%) Age 9 years 58 (15.1%) 58 (44.3%) - - 10 years 112 (29.2%) 67 (51.1%) 44 (34.1%) 1 (0.8%) 11 years 141 (36.7%) 6 (4.6%) 79 (61.2%) 56 (45.2%) 12 years 70 (18.2%) - 6 (4.7%) 64 (51.6%) 13 years 3 (0.8%) - - 3 (2.4%) Language background** Frisian 205 (53.5%) 75 (57.3%) 71 (55.5%) 59 (47.6%) Dutch 139 (36.3%) 42 (32.1%) 47 (36.7%) 50 (40.3%) Regional11 3 (0.8%) - 1 (0.8%) 2 (1.6%) Foreign12 4 (1.0%) 2 (1.5%) - 2 (1.6%) Multilingual 32 (8.4%) 12 (9.2%) 9 (7.0%) 11 (8.9%)
* Of three students the general background information is not known, one of group 6 and two of group 7. Also from one of the students from group 7 the language background is not known, therefore N = 383 and N = 128 applies there ** Based on language spoken with parents
In this research, it is investigated if there are differences between students on Frisian reading comprehension within the different school types and with different language backgrounds. Some decisions were made before the analyses of these questions could start. In Table 3, it can be found that there are only three and four students with a regional and foreign language background. Since these groups are so small, and therefore generalization is not possible, it was decided to remove these students from the sample. Furthermore, the decision was made to remove the students who did not provide any general background information, since they cannot be placed in any group for further investigations. These decisions decreased the total amount of participants to N = 376.
In addition, it was also investigated whether there are differences between students who deal with dyslexia and students without dyslexia, as dyslexia has influence on reading comprehension (Ghesquière & Ruijssenaars, 2008; see chapter 1.1.3 Points of consideration). A
Mann-Whitney U-test revealed a statistical significant difference in the levels on the Frisian
reading comprehension test between children with dyslexia (n = 24, Md = 4, ranking score = 268.69) and children without (n = 352, Md = 3, ranking score = 183.03), U = 2299.5, z = -3.827, p < 0.000, r = 0.20. This indicates that students with dyslexia obtain significantly lower scores on the Frisian reading comprehension tests compared to students without dyslexia. It was therefore decided to remove the students with dyslexia and only investigate students without dyslexia. This decision decreased the total amount of participants to N = 352.
Furthermore, it was analysed whether there are differences between boys and girls, since girls sometimes retain higher scores on reading (Logan & Johnston, 2009; see chapter 1.1.3 Points of consideration). A Mann-Whitney U-test showed that there was no statistical significant difference in the levels on the Frisian reading comprehension test between boys (n = 177, Md = 3) and girls (n = 175, Md = 2) without dyslexia, U = 13948.5, z = -1.654, p = 0.098, r = 0.09. The decision was therefore made to keep boys and girls together as one general group.
11
The regional language background can be defined as a language that is spoken but has no official status in the province, such as Bildts, Stadsfries and Stellingwerfs. The exact regional language of the participants is not known, since there was no option in the questionnaire to describe this.
12
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism With the above decisions made, the following sample was set up, with two regular schools, five schools with Start variety 1, one school with Start variety 2, 12 schools with a Basic trilingual school certificate and two schools with a Plus trilingual school certificate (see Table 4): Table 4: General background information (school type, gender, age, language background) of the complete sample of this master thesis and per separate primary school group
Total (N = 352) Group 6 (N = 120) Group 7 (N = 117) Group 8 (N = 115)
School type None 3TS 27 (7.7%) 23 (19.2%) 2 (1.7%) 2 (1.7%) Start var. 1 81 (23.0%) 9 (7.5%) 30 (25.6%) 42 (36.5%) Start var. 2 20 (5.7%) 0 (0.0%) 9 (7.7%) 11 (9.6%) 3TS Basic 196 (55.7%) 77 (64.2%) 66 (56.4%) 53 (46.1%) 3TS Plus 28 (8.0%) 11 (9.2%) 10 (8.5%) 7 (6.1%) Gender Boy 177 (50.3%) 62 (51.7%) 66 (56.4%) 49 (42.6%) Girl 175 (49.7%) 58 (48.3%) 51 (43.6%) 66 (57.4%) Age 9 years 54 (15.3%) 54 (45.0%) - - 10 years 101 (28.7%) 61 (50.8%) 39 (33.3%) 1 (0.9%) 11 years 127 (36.1%) 5 (4.2%) 73 (62.4%) 49 (42.6%) 12 years 67 (19.0%) - 5 (4.3%) 62 (53.9%) 13 years 3 (0.9%) - - 3 (2.6%) Language background* Frisian 188 (53.4%) 67 (55.8%) 65 (55.6%) 56 (48.7%) Dutch 133 (37.8%) 41 (34.2%) 44 (37.6%) 48 (41.7%) Multilingual 31 (8.8%) 12 (10.0%) 8 (6.8%) 11 (9.6%)
* Based on language spoken with parents, with at least one parent speaking Frisian or Dutch
3.2
M
ATERIALS
All the participants were tested on their Frisian reading comprehension skills with new Frisian reading comprehension tests13 developed by the lectureship FMOO (Varkevisser & Kingma, 2015). Results of the participants were presented in five levels based on norm scales (see Appendix IV). The Frisian reading comprehension tests used in this study were developed as digital tests. The only possibility to use them was via a computer or Ipad connected to the internet. The tests used in this study are pilot tests, since no other norm-referenced Frisian reading comprehension tests are available at the time doing this research. It was important to use these norm-referenced tests because they are method-independent and give the possibility to compare students with each other. See Varkevisser (2015)14 for a justification of how these tests were established.
Before the participants started with the real test, they were asked to fill in a digital questionnaire for the gathering of the general background information. This small questionnaire consisted of 11 questions in total, three open and eight closed. The open questions asked for the name, school and age of the participant; the closed questions asked for the group {6, 7, 8}, gender {boy, girl}, language spoken with mother, father and siblings {Frisian, Dutch, Regional, Foreign}, joy in reading Dutch and Frisian {yes, no, a little} and dyslexia {yes, no}. The questions and answer possibilities were set up bilingual; showing Frisian first followed by Dutch. A paper example of the questionnaire can be found in Appendix III.
After filling in the questionnaire, the participants were asked to fill in the questions that
13
These tests are property of the lectureship FMOO and can therefore not be published in this master thesis. Access to the tests can be gained by contacting the lectureship (lector: dr. A.M.J. Riemersma, a.m.j.riemersma@nhl.nl).
14
Nynke Anna Varkevisser (2775328) Master Thesis Multilingualism belonged to three texts as part of the real test. Each test started with a fictional text, followed by an informative text and finished with an interview text15. A screenshot of one of the texts and one question is given in Appendix V, which are comparable in layout for all the other texts and questions used. All these texts were completely in Frisian; there were no Dutch translations given. Each test consisted of 27 multiple-choice questions with four answer possibilities. Only one answer was correct, the other three were distractors. The questions and answer possibilities were bilingual; first in Frisian, followed by Dutch. The answer possibilities were ordered alphabetically or logically in case numbers were involved. For group 6 and 7 the 27 questions were evenly spread across the test. This means that there were nine questions per text (group 6:
α = 0.73; group 7: α = 0.81). For group 8 the first text consisted of ten questions, the second of
nine and the last text had eight questions (α = 0.79). All questions came in reading order. This means that at first questions about the beginning of the text were asked and at the end questions about the last part of the text were asked. The average time for each test, including the questionnaire, was estimated at 45 to 50 minutes.
3.3
P
ROCEDURE
All the 22 schools that have participated in this study were asked via e-mail to participate16. After they confirmed to participate and informed the lectureship FMOO about the number of students that would participate, they received an e-mail with personal login codes and general passwords. Furthermore, they received a letter with instructions about how to login on the website, how to fill in the questions of each text and how to logoff. The participating schools were asked to let their students take the test in the period March 30th, 2015 until April 10th, 2015. This period was extended with a week since some necessary updates in the system were required.
When the students logged in on the website where the test could be found, they were first asked to fill in the questionnaire. When finished, the students could start the test by clicking on the button below the questionnaire-button; this would lead them to the first text to read. A student had to mark that (s)he had read the text, before the questions relating to that text became available. Only one question at a time was showed; the student had to give an answer before (s)he could go to the next question. When finished with all the questions of the text, a ready button became available. After clicking this, the students could continue to the second text and finally the third text. Teachers were asked to supervise during the test procedure.
The answers students filled in became available in a CSV document per part of the test. This means that per group four CSV documents were obtained, one for the answers on the questionnaire and one for each text. The answers on the closed questions of the questionnaire were processed and recoded into numbers (see Appendix III). In addition, the questions on the texts were recoded into A, B, C, D, for the first, second, third and fourth answer possibility. Furthermore, a correct answer was recoded into 1 and a wrong answer was recoded into 0.
The amount of correct answers was compared to the norm scale dependant of the group (see Appendix IV), which resulted in a letter grade for each participant. These letter grades were then recoded into numbers, with an A-level corresponding to 1, a B-level corresponds to 2, a C-level corresponds to 3, a D-level corresponds to 4, and an E-level corresponds to 5. Consequently, this
15
A fictional text deals with imagined reality; an informative text deals with actual reality; an interview text deals with the actual reality in which opinions about the topic are given.
16
