The origins of teacher multilingualism beliefs
An exploratory study on the development of Dutch primary school teachers’
beliefs on multilingualism
University of Amsterdam
Faculty of Social Sciences
MSc Pedagogical Sciences: Youth at Risk
Samar Hashish (12370002)
Thesis supervisor: dr. Orhan Agirdag
Second reader: prof. dr. Peter de Jong
Study year: 2019-2020
Preface
In the name of the most Merciful and most Forgiving
Dear reader,
Before you lies the conclusion to my journey as a master student at the
University of Amsterdam. It has taken a bit longer than anticipated but I finally
finished it. I have had a couple of setbacks, but I am proud of the final product.
This was my personal attempt to contribute to social justice in the Dutch
education system, especially for those often-marginalized minority groups.
Minority children often find themselves at a disadvantage before they even start
their academic journey and I believe their multilingual abilities are a powerful
way of catching up.
I would like to thank my supervisor, dr. Agirdag, for his patience, understanding
and guidance during my research. I believe that the right supervisor is key to
academic success and for this I am grateful. And finally, I would also like to
thank my family and friends for their constant love and support during this time
and always.
Abstract
The aim of this research was to map out Dutch primary school teachers’ beliefs
on multilingualism and to identify the factors which may contribute to the
development of these beliefs. Three constructs were tested in their ability to
predict teacher beliefs on multilingualism: (1) teacher education, (2) personal
lingual experience and (3) contextual factors. Dutch primary school teachers
(N=180) were asked to fill in an online survey. Of the three constructs, only
contextual factors were found to be significant predictors for teachers’ beliefs
on multilingualism. Especially the schools’ policy on multilingualism seems to
be an important factor: the more accommodating the policies, the more positive
the teachers’ beliefs on multilingualism. One implication of these findings could
be that teachers’ beliefs are not developed based on early experiences and
factual knowledge about multilingualism but rather could be depending on the
context of the school where the teachers work.
Table of contents 1. Introduction ... 5 2. Multilingualism ... 7 2.1 Definitions... 7 2.2 Dimensions ... 7 2.3 Cognitive impact ... 8 3. Teacher beliefs ... 9 3.1 Definitions... 9
3.2 The development of teachers’ beliefs ... 10
3.3 Teachers’ beliefs on multilingualism ... 13
4. This study ... 14
4.1 The conceptual framework... 14
5. Method ... 17
5.1 Participants and sample ... 17
5.2 Procedure ... 17
5.3 Data analysis ... 18
5.4 Variables ... 18
6. Results ... 22
6.1 Research question 1: Multilingual beliefs ... 22
6.2 Research question 2: The constructs ... 23
7. Discussion ... 26
7.1 Theoretical implications ... 27
7.2 Practical implications ... 28
7.3 Limitations and strengths ... 28
7.4 Further research ... 29
7.5 Conclusion ... 29
8. References ... 31
9. Appendices ... 35
9.1 Appendix A: The questionnaire ... 35
1. Introduction
Over the course of decades, the Netherlands has shaped itself to be a country filled with a variety of cultures, ethnicities and languages. Almost 2.5 million citizens in the Netherlands were raised with at least one other language next to Dutch (KNAW, 2018). 22.6 percent of the Dutch citizens were not born in the Netherlands or have at least one parent born outside of the Netherlands (CBS, 2018). This cultural and ethnic diversity results in a linguistic diversity; many immigrants speak the language of their home countries and teach it to their children. Although there are no official numbers on the variety of languages which Dutch citizens speak, based on the background of these people we can assume that many speak an extra language such as Turkish, Arabic, Berber (Moroccan) (KNAW, 2018). Due to e.g. migration, parents of non-Dutch origins often make the choice to speak to their children in their own native language which results in a transgenerational transmission of many foreign languages. Consequently, the classroom becomes an increasingly culturally and linguistically diverse context.
With this increase in cultural and linguistic diversity in the classroom, one might expect that the use of foreign languages within the classroom would increase as well.
However, the opposite is true; the use of foreign languages within the Dutch schools seems to decrease (KNAW, 2018). School policies in the Netherlands are predominantly monolingual even though the Dutch laws allow the existence of multilingual education. In fact, teachers, according to the SLO (2018), seem to believe that facilitating languages other than Dutch withholds students from properly learning the dominant language. This does not completely align with current research on the effects of multilingualism on language acquisition, which shows that maintaining the home language has proven to be beneficial for the development of children’s executive functioning (Bialystok, 2015). For example, bilingual Turkish-Dutch children showed cognitive gains in both verbal working and visuospatial memory tests, even
after correcting for SES and vocabulary (Blom, Küntay, Messer, Verhagen, & Leseman, 2014). One might wonder about the way teachers’ beliefs on multilingualism come to be. Therefore, the aim of this study is to map out the current beliefs which Dutch primary school teachers hold on multilingualism and to identify the factors that may influence these beliefs.
2. Multilingualism 2.1 Definitions
In order to be able to explore the different factors that may influence teachers’ beliefs on multilingualism, it is relevant to first explain this concept. The EU Commission (2007) defines multilingualism as the ability of societies, institutions, groups, and individuals to engage, on a regular basis, with more than one language in their daily lives. In this case, language is defined as a means of communication which includes regional languages, dialects, and sign languages. With regards to the linguistic capabilities, the term
multilingualism applies to competence ranging from partial skill competence to full literacy.
2.2 Dimensions
Cenoz (2013) discusses three dimensions of multilingualism; the individual-social dimension, the proficiency-use dimension, and the bilingualism-multilingualism dimension. In the individual-social dimension one can classify multilingualism as an individual skill but it may also be the use of various languages in a society. Individual and societal levels are not mutually exclusive but may coexist. When looking into individual multilingualism, an important aspect is factored into the analysis: the language proficiency. Some scholars require a maximal proficiency of the language in order to classify someone as multilingual, while other scholars require minimal proficiency. Furthermore, the difference between bilingualism and multilingualism has not always been clear and the terms are often used interchangeably. Some studies may use bilingualism as a generic term, some studies use multilingualism as the generic term (Cenoz, 2013). Depending on the study, the term
bilingualism refers to the use of two language whereas multilingualism or was used to refer to the use of three or more languages. In this thesis I will be using the term multilingualism as a generic term for the use of at least two or more languages.
2.3 Cognitive impact
The cognitive impact of multilingualism has been fairly researched over the past decades. Some research suggest that multilingualism may not always be beneficial to children. For example, it was found that prematurely born multilingual children showed significantly lower cognitive outcomes in comparison to their monolingual peers (Van Veen et al., 2019). Other research has suggested positive effects of mastering multiple languages on the cognitive outcomes of a child, specifically on the executive control (Bialystok, 2011). One explanation for the positive outcomes of multilingualism is that because of the active use of multiple languages simultaneously, the executive control system is recruited into the process of language processing. This is something monolinguals do not have. Since the executive control system is now more frequently used in daily language processing, this system becomes strengthened (Bialystok, 2011). For example, bilingual Turkish-Dutch children showed cognitive gains in verbal working and visuospatial memory tests, even after correcting for SES and vocabulary (Blom, Küntay, Messer, Verhagen, & Leseman, 2014). Additionally, there has been more research implying the effects of multilingualism on the delay of memory loss in the form of dementia and Alzheimer’s disease (Bialystok, Craik, & Freedman, 2007; Craik, Bialystok, & Freedman, 2010; Bialystok, 2011).
3. Teacher beliefs 3.1 Definitions
In the teaching process, two major domains can be identified: (1) the teachers’ thought processes, and (2) teachers’ observable actions (Fang, 1996). These thought processes are also referred to as teacher cognition and involve what teachers know, think, and believe about teaching, students, and content (Borg, 2003; Fang, 1996; Kagan, 1990). The definition of teacher beliefs varies across studies. To define teacher beliefs, two approaches can be used (Fives & Buehl, 2012). The first one includes identifying the content of these beliefs. In the literature we can find six distinct domains that have been examined up until now in teacher beliefs research. These are beliefs about the self, the context or environment, content or knowledge, specific teaching practices, teaching approach and the students. The second approach is to define the construct itself. Conceptualizing the definition of beliefs has proven to be a challenge throughout literature (Hutner, & Markman, 2016). In educational
philosophy literature, a belief is consistently defined as a proposition or a statement of relation among things accepted to be true (Eisenhart, Shrum, Harding, & Cuthbert, 1988; Richardson, 1996 cited in Kyles & Olafson, 2008). It is a mental representation which influences the practice of a teacher once it’s active in their cognition (Hutner, & Markman, 2016).
Specifically, a belief is categorized as a mediating mental representation (Hutner, & Markman, 2016). A mediating mental representation mediates between environmental information, goals, thoughts, and actions. To illustrate this process adequately, I will use an example given by Landsman (2004) as described by Hutner and Markman (2016). A teacher asks a difficult question and calls on white students to answer the question. However, in case of the easy question which “anyone could answer” the teacher turns to black and Latinx students. In this case the goal of the teacher remains the same; he wants a question to be
answered. The environment of the teacher also remains unchanged; the ethnic composition of the classroom did not change with the question. It is the cognitive understanding, the belief, that mediated between the goal and environment and which resulted in different behaviour of the teacher. Beliefs often dictate practice, which is why it is relevant to try and determine what factors may influence the way beliefs are formed.
3.2 The development of teachers’ beliefs
Borg (2003) described a framework to conceptualize the development of teacher cognition. Whereas the term teacher cognition is broad and includes everything -education related- a teacher knows, thinks, and believes, teacher beliefs only refer to whether teachers accept or reject certain notions (Ertmer, 2005). Because teacher beliefs are part of teacher cognition, the framework on the development of teacher cognition by Borg (2003) provides a good theoretical basis to examine the possible factors that may play a role in the development of beliefs. The framework is largely used for the theoretical model in this study. In the
framework, four constructs are discussed. These constructs are (1) schooling, (2) professional coursework, (3) contextual factors and (4) classroom practice and will be elaborated on in the following paragraphs.
3.2.1 Schooling and teachers’ beliefs. Schooling refers to the experience teachers had
during their time as learners themselves and the way they observed their own teachers (Borg, 2003). These early experiences are children’s first encounter with training and learning. In the context of language learning, early experiences with language may contribute to the perception future teachers will have about language learning. If children are told to only speak Dutch, they may proceed to demand the same thing from their students in the future (Öztürk & Gürbüz, 2015).
Not only past language learning experiences in the classroom, but also the teachers’ own experience with multilingualism may play an important role in the development of teachers’ beliefs on multilingualism (Ellis, 2004). Teachers draw upon their own experience with language to fuel their beliefs and conceptions about language learning. Teachers who were multilingual students themselves in the classroom understand their multilingual students because they share the same experiences. Therefore, teachers’ experiences may influence their beliefs and practice. Prior language learning experience is therefore an important factor in the development of teacher beliefs. In this research, prior language learning experience will be operationalized as the experience teachers have had with language in their personal life. This includes for example whether they themselves are multilingual.
3.2.2 Professional coursework and teachers’ beliefs. Professional coursework refers
to the teacher education (in The Netherlands referred to as PABO). Prior to entering the work field, teachers go through extensive programmes in which they are provided with knowledge on the subject of teaching. Knowledge is an important source for teachers’ beliefs (Buehl & Fives, 2009). When teachers were asked which sources of knowledge contributed to their beliefs, they reported six sources: Formal education, formal bodies of education such as books and literature, observational learning, collaboration with others, enactive experiences, and self-reflection. Most of these are essential elements in teachers’ education. A different
line of research suggests that teachers’ beliefs do not necessarily change as easily during their training (Richardson, 2003). This may be the result of a disconnect between the academic sphere and practice, the lack of experience as preservice teachers to understand the
applicability of theory or the short period in which they actually engage with the program. It is therefore key to understand to what extent teacher education provides knowledge and to what extent this knowledge contributes to the teachers’ beliefs within the Dutch context. Teachers have for example stated that they are insufficiently prepared during their training to deal with multilingualism in the classroom (SLO, 2018). Education programmes may have provided future teachers with unrealistic/inappropriate/naïve or insufficient understandings of teaching and learning.
3.2.3 Contextual factors, classroom practice and teachers’ beliefs. Contextual factors
refer to the context of the social, psychological, and environmental realities within the school. These include school policies, parents’- and principals’ requirements of teachers but also the curriculum of the school and the availability of resources for teachers to use in their
classrooms. One theory concerning these factors is the theory of situated cognition (Putnam & Borko, 2000). Rather than treating cognition as an independent core, situated cognition assumes that the physical and social contexts in which certain activities take place are essential to the learning process and therefore affect teachers’ beliefs.
Contextual factors of the institution, for example the school where teachers work, affect teachers’ performance and beliefs (Jamalzadeh & Shahsavar, 2015). Teachers who work in schools with an evenly distributed ethnic composition of the school student body are more inclined towards the support of multilingualism (Pulinx, van Avermaet & Agirdag, 2017).
3.3 Teachers’ beliefs on multilingualism
Previous research suggests that teachers have a negative view on the bilingualism of their students. Some suggest that speaking Turkish instead of Dutch shows a lack of integration and the will to socialize (Bezcioglu-Göktolga, & Yagmur, 2018). In addition, some teachers suggested that once a child reaches school age, Dutch should be the sole used language at home in order to ensure educational and societal success. Children who speak a minority language such as Turkish experience less appreciation from their teachers in comparison to their German speaking peers (Goriot, Denessen, Bakker, & Droop, 2016). This suggests that teachers not only view bilingualism as something negative, but also distinguish between Western languages and non-Western languages. In the Flemish (Dutch-Belgian) context, teachers overall subscribe to more monolingual beliefs (Pulinx, Van Avermaet, & Agirdag, 2017). This affected the extent in which teachers trusted their students; the stronger teachers subscribed to monolingual beliefs, the less they trusted their students. Additionally, a relationship between the ethnic composition of the schools and monolingual beliefs was found. Teachers who worked at schools with a more evenly distributed ethnic composition had showed the most multilingual beliefs.
4. This study
Given the impactful role of teachers’ beliefs on the performance of students, it is important to identify factors that can influence these teacher beliefs, with a special focus on multilingual teacher beliefs. Therefore, this study aims to answer the research question of this Master thesis: “To what extent do Dutch primary school teachers subscribe to multilingualism beliefs and which factors contribute to the development of these beliefs?” In order to answer this research question, I will examine the role of prior education, early language experiences and contextual factors on a school level.
To answer this research question, there will be two sub questions I am going to explore: 1. To what extent do Dutch primary school teachers subscribe to multilingual beliefs? 2. Which constructs play a role in the development of Dutch primary school teachers’
beliefs on multilingualism?
H1. Teachers’ early language learning experiences relate to their beliefs on multilingualism.
H2. Teachers’ education (PABO) relate to teachers’ beliefs on multilingualism.
H3. Contextual factors relate to teachers’ beliefs on multilingualism.
4.1 The conceptual framework
In order to answer these research questions, the theoretical model by Borg (2003) (Figure 1) was used as the basis for the conceptual model (Figure 2) in this study. In this conceptual model, the four concepts in the framework by Borg were reduced to three: (1) Early language experience, (2) teacher education, and (3) contextual factors. Each of these dimensions was then translated into specific variables which were included in the questionnaire.
The construct Schooling (Borg, 2003) was translated into the construct “Early language experience” in this conceptual model. The reason for this shift is because Early
language experience specifically focuses on the aspects in which young children experience, interact with and learn about language rather than how to be a teacher. Therefore, this construct analyses (1) whether or not teachers were raised multilingually themselves, (2) whether or not teachers live in a direct multilingual environment, and (3) if teachers ever aimed to become multilingual by taking a language course.
The construct Professional Coursework (Borg, 2003) was translated into the construct “Education”. The reason for this change is because that the construct by Borg solely focuses what teachers learn during their studies. However, extra training and education outside of the teachers’ education is not included in this construct even though this may also be a source of factual information about multilingualism. Because this extra type of education is included in this study, the name of the construct is changed to “Education” in this study. This construct analyses (1) whether or not the subject of multilingualism was included during the teachers’ studies and (2) whether or not they received additional training or education on the subject of multilingualism.
The constructs Contextual factors and Classroom Practice (Borg, 2003) were merged into one construct “Contextual factors”. The reason for this merge is because rather than completely looking at classroom dynamics, this study only focuses on the degree to which teachers trust their students. The degree to which they trust their students may affect whether or not teachers would allow different languages to be spoken in the classroom. The construct analyses (1) the lingual composition of the students in the school, (2) to which extent school policy accommodates or discourages multilingualism, (3) the type of school and (4) to what degree the teachers trust their students.
Figure 2: conceptual model Early language experience • Multilingual upbringing • Number of multilingual friends
• Taking a language course
Contextual factors
• School policy • School type
• Teacher attitude towards classroom
• Percentage of multilingual children in school
Education
• Overall experience with multilingualism in teacher education • Extra training on multlingualism
Teacher
Beliefs
5. Method 5.1 Participants and sample
The participants in this research are Dutch teachers of the primary education system (N=180). Most teachers are women (89.4%). The ages of the participants ranged between the ages of 21 and 63 (M=37.77, SD= 11.81). The length of their teaching experience ranged between 0 and 40 years (M= 12.56, SD= 10.62). The largest groups of teachers received their education in the province of South-Holland (24.4%) and North-Brabant (20.6%). There were no respondents from the province of Limburg. The teachers mostly worked in public schools (37.8%) and catholic schools (26.7%).
5.2 Procedure
This study is of quantitative nature. The constructs are tested through an online
questionnaire in qualtrics. Due to the unusual circumstances, the fact that primary schools are now (March/April 2020) are closed due to the Corona virus, the initial methodology of this thesis is unachievable. Initially, schools were going to be contacted and asked to fill in a profile of the school after which the teachers in the school would be asked to fill in the questionnaire. Instead, the choice has been made to contact teachers directly via social media through special Facebook groups for Dutch primary school teachers such as “Leerkrachten
basisschool” and “basisonderwijs” and Linkedin to find respondents for the online
questionnaire. Additionally, ten school organisations have been contacted and asked to distribute the questionnaire through their newspaper. And finally, the ‘Gelijke Kansen Alliantie’, an initiative for equal chances in education from the Ministry of Education, was asked to distribute the questionnaire within their primary schoolteachers-network.
Before teachers start the questionnaire, the teachers were asked to give consent for the use of the data they provided for this research.
5.3 Data analysis
The quantitative data in this research consists of the responses of teachers to an online survey. This data was handled with SPSS 24. The sample included no missings and was tested
positively for the assumptions of homoscedasticity, linearity, multivariate normality and multicollinearity, associated with multiple regression analyses (Appendix B). With respect to the hypotheses, for each of the three constructs, a multiple regression analysis was conducted in order to assess whether the constructs by themselves had influence on the teachers’ beliefs on multilingual. Additionally, a final model was added to assess the collective influence of these constructs together on the teachers’ beliefs on multilingualism.
5.4 Variables
The variables factored into the analyses are all subscribed to one of the three constructs which are tested in this study (See table 2 for descriptives). School type and percentage of multilingual students were added as control variables.
5.4.1 Multilingual beliefs. This construct measures the teacher’s perception of
children’s multilingual behaviour and their beliefs on this matter. This construct is measured by eight items such as “Non-Dutch speaking pupils should not be allowed to speak their
home language with each other on the playground”, inspired by Pulinx, Van Avermaet, and
Agirdag (2017). Teachers respond to these statements on a 4-point likert-scale from “Completely disagree” to “Completely agree”. Items 1 through 4 were recoded and the mean score was then computed. The construct reliability, a Cronbach’s alpha of α = .782 was found. A full overview of the items and the percentage of teachers who agreed with them can be found in Table 2.
5.4.2 Trust in students. The class context measures the degree to which a teacher
trusts their students. This is measured through 4 statements such as “My students are
disagree” to “Completely agree”. For the reliability, a Cronbach’s alpha of α = .606 was
found. (M=3.06, SD=.37)
5.4.3 Multilingualism education. This construct measures the degree to which
teachers have received substantive knowledge on multilingualism or practical knowledge on how to deal with multilingualism as a teacher. Teachers are also asked whether they received additional training on the subject outside the regular curriculum. For the reliability, a
Cronbach’s alpha of α = .714 was found. However, the final item on additional schooling
was deleted to gain a Cronbach’s alpha of α= .868. The item was later added in as an additional variable. (M=2.19, SD=.71)
5.4.4 School policy. This is measured through three statements such as “In our school, non-Dutch speaking students are not allowed to speak in their native language to one another”. Teachers respond to these statements with “Yes” or “No”. For the reliability, a
Cronbach’s alpha of α = .590 was found. (M=.48, SD=.33)
5.4.5 Percentage of multilingual students. The percentage of multilingual students
in schools was measured through an estimation made by the teachers (0-100) of multilingual students in the school they work at. The mean percentage teachers reported was 49 percent.
5.4.6 Multilingual environment. To measure the teachers’ multilingual environment,
teachers were asked how many people in their current personal environment are multilingual. They could answer on a scale from no person (score=1) to more than 3 people (score=5). (M=3.49, SD=.1.70)
5.4.7 Additional schooling. In order to measure additional schooling teachers were
asked how often they received additional schooling on the subject of multilingualism next to their regular teachers’ education. They could answer on a scale from no schooling (score=1) to more than 2 times (score=4) (M=1.83, SD=1.14).
5.4.8 Multilingual upbringing. In order to measure teachers’ personal experience
with multilingualism, teachers were asked in which language they were raised. They could choose between three options; raised with the Dutch language, raised with a non-Dutch language, or raised with both the Dutch and a non-Dutch language. Because all teachers spoke the Dutch language, the variable was recoded into two options in which the last two categories were taken together as multilingual (scored=1) and the first category as non-multilingual (scored=0). 33.3 percent of the respondents was non-multilingual. Table 1 describes the different languages which were spoken by the multilingual teachers.
Table 1: Distribution of the languages spoken by multilingual teachers
Language %
Arabic 33.3
Turkish 13.3
English 6.7
Berbers (Moroccan dialect) 18.3
Dutch dialect/Fries 8.3
Other 20.0
5.4.9 Language course. In order to measure teachers’ interest in multilingualism,
they were asked whether they ever did a language course. They could answer with yes
(score=1) and no (score=0). 38.9 percent of the respondents reported to have done a language course.
5.4.10 School type. In order to measure the teachers’ work context teachers were
asked what type of school they work at. The most popular school types were provided as options for the teachers. There options were Public schools (score=1, 37.8%), Catholic schools (score=2, 26.7%), Protestant schools (score=3, 14.4%), Islamic schools (score=4, 13.3%) and other type of schools (score=5, 7.8%).
Table 2: Descriptive statistics of the sample (Percentages provided for categorical variables and Cronbach’s Alpha for computed constructs)
N Min. Max. M SD Alpha
Multilingual Beliefs 180 1 4 2.43 .48 .782
Trust in students 180 1 4 3.06 .37 .606
Multilingualism education 180 1 4 2.19 .71 .868
School lingual policy 180 0 1 .48 .33 .590
Percentage of multilingual students 180 0 100 49.43 35.60 Multilingual environment 180 1 5 3.49 1.70 Additional schooling 180 1 4 1.83 1.14 Multilingual upbringing 180 0 1 33.3% Language course 180 0 1 38.9% School type 180 Islamic 24 0 1 13.3% Public 68 0 1 37.8% Catholic 48 0 1 26.7% Protestant 26 0 1 14.4% Other 14 0 1 7.8%
6. Results
The goal of this study is to determine which factors can successfully predict teacher beliefs on multilingualism. The research question was “To what extent do Dutch primary school teachers subscribe to multilingualism beliefs and which factors contribute to the development of these beliefs?” The first research question concerned the teachers’ beliefs on multilingualism and the second one concerned the conceptual model. There were three hypotheses, based on the threesss constructs in the testing model of this research (Figure 2). Table 3 provides an overview of all the conducted regression analyses which will be
elaborated on in this chapter.
6.1 Research question 1: Multilingual beliefs
Table 3 provides an overview of the extent to which teachers agreed to the statements on multilingual beliefs (N=180). The responses ‘agree’ and ‘completely agree’ were both
considered in these percentages. About 55 percent of the teachers agreed that children should not be allowed to speak their home language in school (item 1). The majority of the teachers agreed that allowing students to speak their home language with each other would result in the exclusion of others (item 4). A little over half of the teachers (55.3%) agreed that children should be able to help each other in their home language (item 8). Overall, teachers do not seem to be inclined towards facilitating multilingualism in school.
Table 3: Items of Multilingual Beliefs Scale
Item Description (Completely)
agree
1 Non-Dutch speaking pupils should not be allowed to speak their home language with each other on the playground
55% 2 Non-Dutch speaking pupils should not be allowed to speak their
home language at school
75% 3 It is in the interest of the pupils when they are punished for
speaking their home language at school
2.8% 4 If non-Dutch speaking students speak their home language with
each other, other students will become excluded
5 The school library (classroom library, media library) should also include books in the different home languages of the pupils
55.6% 6 Non-Dutch speaking pupils should be offered the opportunity to
learn their home language at school
25.5% 7 Non-Dutch speaking pupils should be offered regular subjects in
their home language
6.1% 8 Non-Dutch speaking pupils should be allowed to help each other
in their home language
53.8%
6.2 Research question 2: The constructs
6.2.1 Multilingual beliefs and experience. In order to test the hypothesis that
teachers’ personal experience with language affects teachers’ beliefs on multilingual, I conducted a multiple regression analysis with the three variables on experience. Overall, the model did not explain a large or significant proportion of the variance in teachers’ beliefs on multilingual (R2=.02, F(3)= .107, p= .956). Of the three predictors, none were found
significant. Teachers’ personal interest or experience with multilingualism did not predict neither positive nor negative multilingualism beliefs. The hypothesis is therefore rejected.
6.2.2 Hypothesis 2: Multilingual beliefs and multilingualism education. In order to
test the hypothesis that teachers’ education affects teachers’ beliefs on multilingualism, I conducted a regression analysis with the computed construct on teachers’ education and an additional variable on whether or not teachers received additional schooling on
multilingualism. Overall, the model explained a small, but statistically significant proportion of the variance in multilingual beliefs (R2=.06, F(2)= .107, p= .040). When teachers attended additional schoolings on multilingualism, they rate multilingualism slightly more positive (b=.102, SE=.031, p=.001). The overall inclusion of multilingualism in teachers’ education did not have any statistical significance in the model (b=-.049, SE=.050, p=.332). Therefore, the hypothesis that teachers’ education influences multilingualism beliefs is mostly rejected, only additional education shows some effect.
6.2.3 Hypothesis 3: Multilingualism beliefs and contextual factors. In order to test
the hypothesis that school context affects teachers’ beliefs on multilingualism, I conducted a regression analysis with the variables school policy, the dummy variables of school type, teacher’s general attitude towards their class and the percentage of multilingual children in school. Overall, this model proved to be the strongest of the three hypotheses. The model explained a quite large and significant proportion of the variance in the teachers’ beliefs on multilingualism (R2=.41, F(7)= 16.915, p< .001). In this model, three significant variables were found for the variance in teachers’ beliefs on multilingualism: the percentage of multilingual students (b=.004, SE=.001, p<.001), school lingual policy (b=.805, SE=.089,
p<.001) and school type, specifically Islamic schools (b=-.355, SE=.143, p<.001). The first
two variables show positive values. For the percentage of multilingual students this means that if teachers work in schools with a higher percentage of these students, they show a slight but significant increase in positive multilingual beliefs. The same is found for school policy, more statistically more positive multilingual beliefs were observed in teachers who work in schools with a more multilingualism accommodating school policy. Figure 3 provides an overview of the mean policies for each type of school. Whether the teacher trusted their students was not statistically relevant. The hypothesis that the context the teacher works in influences teachers’ beliefs on multilingualism is therefore partly accepted.
Figure 3: Mean Multilingual Policies per School Type 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Public Schools Catholic Schools Protestant Schools Islamic schools Other Schools Multilingualism accomodating policy
6.2.4 The final multiple regression model. To answer the research question “which
factors influence teachers’ beliefs on multilingualism?” I conducted a multiple regression analysis which included all three constructs in the model. The model explained a quite large and significant proportion of the variance in teachers’ beliefs on multilingualism (R2=.43,
F(12)= 10.477, p < .001) with 3 significant predictors: School policy (b=.814, SE= .092 p
<.001) and school type (Islamic) (b= -.355, SE=.143, p=.014) and percentage of multilingual students (b= .003, SE=.001 p=.018). These statistics are completely in line with the third hypothesis on contextual factors. The other two constructs, education, and experience /proved to be nonsignificant. Even additional schooling, which was significant in the second
hypothesis, became nonsignificant in the final model (b=.047, SE=.028, p=.101).
Table 3: Regression models Personal experience Contextual factors Teacher education Final model B SE p B SE p B SE p B SE p Multilingual upbringing .001 .086 .117 .076 Language course .042 .077 -.021 .061 Multilingual environment -.005 .024 .000 .020 Trust in students .094 .080 .057 .081 Percentage of multilingual students .004 .001 *** .003 .001 ** School policy School type .805 .089 *** .814 .092 *** Type: Islamic -.303 .140 ** -.355 .143 * Type: Public .028 .115 -.007 .116 Type: Catholic -.022 .120 -.028 .120 Type: Protestant Ref: Others -.066 .130 -.099 .133 Multilingualism education -.049 .050 -.059 .041 Additional schooling .102 .031 ** .047 .028 *p < .05 **p < .01 ***p < .001
7. Discussion
This study aimed to (1) map the degree to which Dutch primary school teachers subscribe to multilingualism beliefs and (2) determine which factors contribute to the development of these beliefs. The research question was “To what extent do Dutch primary school teachers subscribe to multilingualism beliefs and which factors contribute to the development of these beliefs?”. Three constructs, based on Borg’s (2003) framework on development of teacher cognition, were tested to answer the research question. These constructs were (1) teachers’ personal language experience, (2) teachers’ education and (3) contextual factors (Figure 4).
Figure 4: Conceptual model with its significance
Experience
• Multilingual upbringing • Number of multilingual
friends
• taking a language course
Contextual factors
• School policy* • School type*
• Teacher attitude towards classroom
• Percentage of multilingual children in school*
Education
• Overall experience with multilingualism in teacher education • Extra training on multlingualism
Teacher
Beliefs
7.1 Theoretical implications
Based on the conducted analyses and contrary to the expectations, two of the three tested constructs did not show any statistically significant relationships with teachers’ beliefs on multilingualism (Figure 4). Teachers’ beliefs on multilingualism did not significantly differ between teachers with and without multilingual background or if they had any
experience with multilingualism. Whether multilingualism was included in the curriculum of the PABO did not seem to have any effects on the multilingualism beliefs either. Although this result was not expected, it is in line with a review by Kagan (1992), which suggested that teachers’ beliefs did not change during their studies but remained stable. These findings may implicate that teachers’ beliefs on multilingualism do not necessarily rely on factual
knowledge and experiences obtained throughout the teachers’ (educational) life, be they positive or negative.
Contextual factors seemed to be the only factors which are statistically significantly
related to teachers’ beliefs on multilingualism in this study. The estimated percentage of multilingual students (by the teacher) and lingual school policy both showed positive linear relationships with multilingualism beliefs. Teachers who worked at a school with a more accommodating school policy on multilingualism showed more positive inclinations towards multilingualism. Teachers who worked at a school with a higher perceived percentage of multilingual students showed a slightly more positive inclination to multilingualism. There are multiple implications for these results. One implication could be that there is a causal relationship between the variables in which the school context influences the teachers’ beliefs on multilingualism. However, a different implication could be that there is a possible
selection effect. This would mean that, for example, teachers who hold more positive beliefs on multilingualism could be more likely to choose to work at schools with a climate they
identify with, one which encourages multilingualism. The direction of this relationship is unclear.
What is most surprising is that school contextual factors, the only significant factors in this research, are not as common in research on multilingualism beliefs as the
nonsignificant factors. School contextual factors were only briefly mentioned by Borg (2003) in that school context may play some role in the development of teacher beliefs. School policies define the parameters and dictate the rules which teachers have to work with. However, these findings support the notion that cognition can be situated (Putnam & Borko, 2000). The physical context seems to be essential to the development of multilingualism beliefs.
7.2 Practical implications
The findings in this research have implications for policy. One of the most important findings in this study is that teachers’ beliefs on multilingualism seem to depend on the context of the schools the teachers work at. In order to promote positive teachers’ beliefs on
multilingualism, school policies could focus on accommodating multilingualism. This would result in the provision of better tools for teachers to accommodate the linguistic diversity in their classrooms.
7.3 Limitations and strengths
One strength of this study is its exploratory and quantitative nature. Not a lot of research on the development of teacher beliefs exists and most of the available studies are of qualitative nature. This study tests a clear conceptual framework with a large quantitative sample. Another strength of this study is the diversity in its respondents. Every Dutch province was represented in the sample which allows for the results to be more generalizable.
There are a few limitations to this study. The first limitation concerns the methodology of the study. It was first intended to target schools rather than individual
teachers. This however proved to be difficult with the current societal situation of the global COVID-19 pandemic. Consequently, data could not be clustered which may have influenced the data. The second limitation of the study concerns the distribution of the data sample. A third of the teachers in the sample were multilingual which may not be representative of the general population. Additionally, the sample included a large proportion of teachers who work at an Islamic school. This is not an accurate representation of the Dutch context in which less than one percent on the primary schools is of the Islamic faith. Therefore, the results of this study may not be completely generalizable.
7.4 Further research
This study provides groundwork for future research. Because of the explorative nature of this study, the relationship between contextual factors and teachers’ beliefs on multilingual only scratched the surface. The relationship between teachers’ beliefs on multilingualism and school lingual policy is one that requires more investigation to determine whether the relationship could be causal. Additionally, further research should be done to define the contextual factors and to investigate its relationship with multilingual beliefs. Finally, this study solely focused on the relationship between predictors and teacher beliefs but did not consider the practice consequents of these beliefs. Therefore, I recommend that future research also focuses on the consequents of these teacher beliefs on practice.
7.5 Conclusion
The research question in this thesis was “To what extent do Dutch primary school teachers subscribe to multilingualism beliefs and which factors contribute to the development of these beliefs?”. Overall teachers varied significantly in the extent to which they subscribe to multilingualism beliefs and contextual factors such as school lingual policies seemed to be the strongest predictors for the beliefs teachers hold on their students’ multilingualism
however the causality of this relationship is debatable. Further research is required to
8. References
Bezcioglu-Goktolga, I., & Yagmur, K. (2018). Home language policy of second-generation Turkish families in the Netherlands. Journal of Multilingual and Multicultural
Development, 39, 44-59. https://doi.org/10.1080/01434632.2017.1310216
Bialystok, E. (2015). Bilingualism and the development of executive function: The role of attention. Child Development Perspectives, 9, 117–121. doi:10.1111/cdep.12116 Bialystok, E., Craik, F. I., & Freedman, M. (2007). Bilingualism as a protection against the
onset of symptoms of dementia. Neuropsychologia, 45, 459-464. https://doi.org/10.1016/j.neuropsychologia.2006.10.009
Bialystok, E. (2011). Reshaping the mind: the benefits of bilingualism. Canadian Journal of
Experimental Psychology/Revue Canadienne de Psychologie Expérimentale, 65,
229-235. https://doi.org/10.1037/a0025406
Blom, E., Küntay, A. C., Messer, M., Verhagen, J., & Leseman, P. (2014). The benefits of being bilingual: Working memory in bilingual Turkish–Dutch children. Journal of
Experimental Child Psychology, 128, 105-119.
https://doi.org/10.1016/j.jecp.2014.06.007
Borg, S. (2003). Teacher cognition in language teaching: A review of research on what language teachers think, know, believe, and do. Language Teaching, 36, 81-109. https://doi.org/10.1017/S0261444803001903
Buehl, M. M., & Fives, H. (2009). Exploring teachers' beliefs about teaching knowledge: Where does it come from? Does it change? The Journal of Experimental Education, 77, 367-408. https://doi.org/10.3200/JEXE.77.4.367-408
CBS (2019). https://opendata.cbs.nl/statline/?dl=107C4#/CBS/nl/dataset/37296ned/table Cenoz, J. (2013). Defining multilingualism. Annual Review of Applied Linguistics, 33, 3-18.
Craik, F. I., Bialystok, E., & Freedman, M. (2010). Delaying the onset of Alzheimer disease: Bilingualism as a form of cognitive reserve. Neurology, 75, 1726-1729.
https://doi.org/10.1212/WNL.0b013e3181fc2a1c
Eisenhart, M. A., Shrum, J. L., Harding, J. R., & Cuthbert, A. M. (1988). Teacher beliefs: Definitions, findings, and directions. Educational policy, 2, 51-70.
Ellis, E. M. (2004). The invisible multilingual teacher: The contribution of language background to Australian ESL teachers' professional knowledge and
beliefs. International Journal of Multilingualism, 1, 90-108. https://doi.org/10.1080/14790710408668181
Ertmer, P. A. (2005). Teacher pedagogical beliefs: The final frontier in our quest for technology integration?. Educational Technology Research and Development, 53, 25-39.
European Commission. (2007). Final report: High level group on multilingualism. Luxembourg: European Communities. Retrieved from
http://biblioteca.esec.pt/cdi/ebooks/docs/High_level_report.pdf
Fang, Z. (1996). A review of research on teacher beliefs and practices. Educational
Research, 38, 47-65. https://doi.org/10.1080/0013188960380104
Fives, H., & Buehl, M. M. (2012). Spring cleaning for the “messy” construct of teachers’ beliefs: What are they? Which have been examined? What can they tell us? APA
educational psychology handbook, 2, 471-499. https://doi.org/10.1037/13274-019
Goriot, C., Denessen, E., Bakker, J., & Droop, M. (2016). Benefits of being bilingual? The relationship between pupils’ perceptions of teachers’ appreciation of their home language and executive functioning. International Journal of Bilingualism, 20, 700-713. https://doi.org/10.1177/1367006915586470
Hendrics, C. (June, 2018). Islamitische basisscholen zoeken naar balans. Accessed from https://www.universiteitleiden.nl/nieuws/2018/06/beemsterboer-islamitische-basisscholen-zoeken-naar-balans
Hutner, T. L., & Markman, A. B. (2016). Proposing an operational definition of science teacher beliefs. Journal of Science Teacher Education, 27, 675-691.
https://doi.org/10.1007/s10972-016-9480-5
Jamalzadeh, M., & Shahsavar, Z. (2015). The effects of contextual factors on teacher’s beliefs and practices. Procedia-Social and Behavioral Sciences, 192, 166-171. doi:10.1016/j.sbspro.2015.06.024
Kagan, D. M. (1990). Ways of evaluating teacher cognition: Inferences concerning the Goldilocks principle. Review of Educational Research, 60, 419-469.
https://doi.org/10.3102/00346543060003419
Kagan, D. M. (1992). Professional growth among preservice and beginning teachers. Review
of Educational Research, 62, 129-169. https://doi.org/10.3102/00346543062002129
KNAW (2018). Talen voor Nederland. Amsterdam: Koninklijke Nederlandse Akademie van Wetenschappen. Accessed from https://www.knaw.nl/nl/actueel/publicaties/talen-voornederland
Kyles, C. R., & Olafson, L. (2008). Uncovering preservice teachers' beliefs about diversity through reflective writing. Urban Education, 43, 500-518.
https://doi.org/10.1177/0042085907304963
Landsman, J. (2004). Confronting the racism of low expectations. Educational Leadership,
62, 28–32.
Öztürk, G., & Gürbüz, N. (2016). The impact of early language learning experiences on EFL teachers’ language teaching beliefs and practices. ELT Research Journal, 5.
Pulinx, R., Van Avermaet, P., & Agirdag, O. (2017). Silencing linguistic diversity: The extent, the determinants and consequences of the monolingual beliefs of Flemish teachers. International Journal of Bilingual Education and Bilingualism, 20, 542-556. https://doi.org/10.1080/13670050.2015.1102860
Putnam, R. T., & Borko, H. (2000). What do new views of knowledge and thinking have to say about research on teacher learning? Educational Researcher, 29, 4-15.
https://doi.org/10.3102/0013189X029001004
Richardson, V. (2003). Preservice teachers’ beliefs. Teacher beliefs and classroom
performance: The impact of teacher education, 6, 1-22.
SLO (2018). Meertaligheid in Primair en Voortgezet Onderwijs accessed from
http://downloads.slo.nl/Repository/meertaligheid-in-primair-en-voortgezet-onderwijs.pdf Van Veen, S., Remmers, S., Aarnoudse‐Moens, C. S. H., Oosterlaan, J., van Kaam, A. H., &
van Wassenaer‐Leemhuis, A. G. (2019). Multilingualism was associated with lower cognitive outcomes in children who were born very and extremely preterm. Acta
9. Appendices 9.1 Appendix A: The questionnaire
Q1
Geslacht
• Vrouw Vrouw Vrouw Vrouw
• Man Man Man Man
• Anders Anders
Q2
Hoe oud ben je?
Q4
Hoelang ben je al docent? (in jaren)
Q3
In welke provincie heb je de PABO gedaan?
• Noord-Holland • Zuid-Holland Zuid-Holland • Zeeland • Noord-Brabant • Utrecht • Flevoland • Friesland • Groningen • Drenthe • Overijssel • Gelderland • Limburg • Anders Anders Block 1 Block Options Q10
Op de school waar ik werk mogen anderstalige kinderen hun thuistaal niet spreken in de klas.
• Waar Waar Waar Waar
• Niet waar
Q8
Op de school waar ik werk mogen anderstalige kinderen hun thuistaal spreken op de speelplaats.
• Waar Waar Waar Waar
• Niet waar
Q7
Op de school waar ik werk mogen anderstalige ouders hun thuistaal spreken op de speelplaats.
• Waar Waar Waar Waar
• Niet waar
Q10
Op de school waar ik werk wordt verwacht van ouders dat ze thuis Nederlands spreken met hun kind.
• Waar Waar Waar Waar
• Niet waar
Q11
De school waar ik werk is een...
• Openbare school
• (Rooms-) Katholieke school
• Protestantse school
• Islamitische school
• Anders Anders
Q11
Het percentage anderstalige leerlingen op de school waar ik werk is ongeveer...
010 20 30 40 50 60 70 80 90 100 Percent age Block 2 Block Options Q15
Helemaal niet
mee eens Niet mee eens Mee eens
Helemaal mee eens Anderstalige leerlingen zouden op de speelplaats onderling geen vreemde taal mogen spreken.
Anderstalige
leerlingen zouden in de klas onderling geen vreemde taal mogen spreken. Om Nederlands te stimuleren zouden anderstalige leerlingen gestraft kunnen worden als ze onderling in een andere taal spreken. Als anderstalige leerlingen andere talen mogen spreken, worden andere leerlingen uitgesloten. Een schoolbibliotheek (klasbibliotheek, mediatheek) moet ook boeken aanbieden in de moedertaal van de leerlingen. Anderstalige leerlingen moeten op school de mogelijkheid krijgen om hun moedertaal te leren. Anderstalige leerlingen moeten reguliere vakken ook aangeboden krijgen in hun moedertaal. Anderstalige
Helemaal niet
mee eens Niet mee eens Mee eens
Helemaal mee eens de klas elkaar mogen
helpen in hun moedertaal. Block 3 Block Options Q24
De volgende vragen gaan over je eigen handelen in de klas:
Helemaal niet
mee eens Niet mee eens Mee eens
Helemaal mee eens Ik sta mijn leerlingen
toe een andere taal dan Nederlands te gebruiken om de leerstof uit te leggen aan een klasgenoot. Ik vraag mijn leerlingen om school gerelateerde materie voor elkaar te vertalen in hun thuistaal.
Ik sta mijn leerlingen toe een andere taal dan Nederlands te spreken in de klas. Ik zorg ervoor dat mijn leerlingen ook boeken in hun thuistaal kunnen lezen in de klas. Block 4 Block Options Q28
De volgende vragen gaan over je eigen leerlingen:
Nooit Bijna nooit Vaak Heel vaak
Mijn leerlingen zijn betrouwbaar. Ik kan erop rekenen dat leerlingen hun werk doen.
Nooit Bijna nooit Vaak Heel vaak Ik moet mijn leerlingen nauwlettend in de gaten houden. Mijn leerlingen spieken of plegen bedrog als ze er de kans toe zien.
Q33
Het percentage anderstalige leerlingen in mijn klas is ongeveer…
010 20 30 40 50 60 70 80 90 100 Percent age Block 5 Block Options Q34 Ik ben opgevoed…
• …met alleen de Nederlandse taal.
• …met de Nederlandse taal en een andere taal.
• …met een andere taal dan het Nederlands.
Q35
Welke taal?
• Arabisch
• Turks Turks Turks Turks
• Surinaams
• Engels Engels
• Anders, namelijk
Q17
Ik heb mensen in mijn directe omgeving die zijn opgevoed met minimaal een niet-Nederlandse taal
• 1 persoon
• 2 personen
• 3 personen
• Meer dan 3 personen
Q18
Heb je ooit een taalcursus gedaan?
• Ja Ja Ja Ja
• Nee Nee Nee Nee
Q19
Welke taal?
Block 6 Block Options Q20
De volgende vragen gaan over je docentopleiding:
Helemaal niet
mee eens Niet mee eens Mee eens
Helemaal mee eens Mijn opleiding heeft
mij de pedagogische vaardigheden
gegeven om te kunnen omgaan met meertaligheid met betrekking tot de prestaties van leerlingen.
Mijn opleiding heeft mij voorbereid om een basiskennis en begrip van
meertaligheid te demonstreren. Mijn opleiding heeft mij genoeg voorbereid om een meertalige groep scholieren te onderwijzen. Q22
In hoeveel vakken is het onderwerp meertaligheid aan bod gekomen?
• 1 vak 1 vak 1 vak 1 vak
• 2 vakken
• Meer dan 2 vakken
Q21
Ik heb naast mijn opleiding extra nascholing gevolgd over meertaligheid
• 0 keer 0 keer
• 1 keer 1 keer
• 2 keer 2 keer
9.2 Appendix B: Statistical assumptions
1. Linear relationship between independent variables and dependent variable
Based on the scatterplot with all the variables for this study, one can conclude that the first assumption is in fact met; all the independent variables show a linear relationship with the dependent variable of teacher beliefs.
2. No multicollinearity
The second assumption is the assumption of no multicollinearity. In order to test for this assumption, I performed a collinearity test with the regression analysis. In order for the assumption to be met, no variable should show a variance increase factors (VIF) of higher than 10 and tolerance no higher than .2. Additionally, no variables should correlate stronger with each other than 0.8. This is met for all variables but two; the percentage of multilingual children in the classroom (tolerance=.103, VIF=9.673) and the percentage of multilingual
children in the school (tolerance=.103, VIF=9.701) correlate strongly (r=.944, p<.001). Therefore, I made the choice to eliminate the variable percentage of multilingual children in the classroom in order to still meet the assumption.
3. Values of the residuals are independent
The third assumption tested for is the assumption that the residuals are independent. In order to test this, the Durbin-Watson test was conducted. A value between 1 and 3 is sufficient for the assumption to be met (Durbin-Watson=1.909)
Model Summaryb Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics Durbin-Watson R Square Change F Change df1 df2 Sig. F Change 1 .655a .429 .388 .38178 .429 10.477 12 167 .000 1.909
a. Predictors: (Constant), Ik heb naast mijn opleiding extra nascholing gevolgd over meertaligheid, Experience_binary, Education_new, Context1=Openbare school, Heb je ooit een taalcursus gedaan?, Classroom, Schoolpolicy_new, Context1=Protestantse school, Ik heb mensen in mijn directe omgeving die zijn opgevoed met minimaal een niet-Nederlandse taal, Context1=Islamitische school, Het percentage anderstalige leerlingen op de school waar ik werk is ongeveer... - Percentage, De school waar ik werk is een...
b. Dependent Variable: Beliefs
4. Variance of the residuals is constant
The fourth assumption to be met is the assumption of homoscedasticity which can be tested through the examination of a plot of the standardized residuals by the regression standardized predicted value. Based on the figure, I conclude that this assumption is met.
The fifth assumption to be met is the assumption of residual normality which can be tested through the examination of a P-plot of the regression model. Based on the figure, I conclude that this assumption is met.
