Differences in English teachers’ beliefs and practices and inequity in Austrian English language education: could plurilingual pedagogies help close the gap?

ABSTRACT Educational research in Austria has demonstrated that early tracking into academic secondary schools and middle schools results in inequities for students with low socioeconomic status, from migration backgrounds and/or with German as an additional language. It has also found gaps in performance between students in the two school types, including in English. To better understand the reasons for this, we collected data from teachers in both school types about their beliefs and classroom practices. Teachers in academic secondary schools indicate higher achievement levels for English amongst their students compared to teachers in middle schools. Descriptive and inferential analyses reveal higher perceived levels of motivation amongst academic secondary school compared to middle school students, and students’ perceived level of motivation influenced English language outcomes in both school types. Moreover, there were significant differences aligning with school type in teachers’ reported language use in the classroom and their perceptions of students’ access to English outside of school. These findings suggest factors that contribute to gaps in achievement in English language education according to school type. The article closes by considering how plurilingual pedagogies might be employed to transform beliefs, alleviate disadvantage and support equity in English language education in Austria and beyond.


Introduction
This article considers equity in English language education in Austria by comparing the beliefs and practices of English teachers in academic secondary schools with those of middle schools to see if they might be contributing to the gaps in learning outcomes between these schools. In the Austrian education system, students are tracked after only four years of comprehensive schooling into one of two different school types: professionally oriented middle schools or the more academically oriented general academic secondary school. Research has consistently established that this system of early tracking results in an inequitable disadvantage for students from migration backgrounds and/or with German as an additional language, and for students with low socioeconomic status, both of whom are over-represented in middle schools (Schreiner et al., 2020;Statistik Austria, 2019). There is a significant gap in performance between students in these two school types in all the core areas of the curriculum, including English, with middle school students consistently performing below the national average. Their peers in academic secondary schools achieve overall much higher results. To better understand why this is the case, we collected data from teachers in both school types on their beliefs about students' motivations and abilities in English language learning, as well as their classroom practices. In this article, we present descriptive and inferential analyses of the data to explore whether the beliefs and practices of teachers at these two different school types differ, particularly with regards to perceived learning outcomes, students' motivations to learn English, students' access to English outside of school and teachers' reported language practices in the classroom. We find that teachers' beliefs align with national assessments in that those teaching in academic secondary school indicate higher achievement levels for English amongst their students compared to those teaching in middle school. Inferential analyses of the findings reveal higher perceived levels of motivation amongst academic secondary school students compared to middle school students, and the perceived level of students' motivation influenced English language outcomes in both school types. Moreover, there were significant differences aligning with school type in teachers' reported language use in the classroom and in their perceptions of students' access to English outside of school. These findings thus suggest factors that contribute to gaps in achievement in English language education according to school type, which also enhance broader social inequalities. To close the article, we consider how plurilingual pedagogies might be employed to transform beliefs, alleviate disadvantage and support equity in English language education in Austria and beyond. Educational equity in the Austrian school system There are two commonly recognised aspects of educational equity. The first is fairness: a person's circumstancese.g. their gender, socioeconomic status, ethnic origin or language backgroundshould not be an obstacle to achieving their educational potential. The second is inclusion: education should ensure a basic minimum standard of education for all (OECD, 2008). A fair and inclusive system makes the advantages of education available to all, and this, in turn, is one of the most powerful means of making society more equitable, and enhancing social cohesion and trust (OECD, 2008). In an equitable school system, all students in all tracks of schooling in Austria would be able to access high-quality English language learning, regardless of their background and the location of their school. However, research has consistently established that there are educational disadvantages for students from migration backgrounds and/or with German as an additional language, and for students with low socioeconomic status, who are over-represented in Austrian middle schools (Bruneforth et al., 2016;Herzog-Punzenberger, 2017;Schreiner et al., 2020;Statistik Austria, 2019). Since parents from a migration background are also more likely to have a lower educational level, their children often face a double disadvantage at school (Oberwimmer et al., 2019, p. 27). There is a significant gap in performance between students in these two school types in all the core areas of the curriculum, including English, with middle school students consistently performing below the national average. Educational outcomes in urban middle schools with high numbers of students from lower socioeconomic backgrounds and who speak languages other than German at home tend to be the lowest (BIFIE, 2020). The 2019 national standards assessments for English found that 73% of middle school students do not meet the desired learning outcomes set in the curriculum (CEFR level B1 or above in all areas of competence), compared to 31% of students at general academic secondary schools. L1 German students outperformed their peers in English by 22-35 points. When socioeconomic status is controlled for, 1 meaning that students with similar socioeconomic backgrounds in the two school types are compared, the difference between these two groups was substantially reduced (BIFIE, 2020, p. 75;Maaz et al., 2008). The question thus arises as to why there is this gap in English language attainment and whether teacher beliefs and practices might explain part of this difference.

English language education
Equitable access to English language education is important because English (along with German and Math) is one of the core areas of the curriculum. This is surely due to the wide-spread recognition that today English has the status as an international language and that most of its usages are by non-native speakers in a number of professional, educational and social contexts (Seidlhofer, 2011). The European Commission (2002), among other official bodies, has repeatedly issued recommendations linking competences in foreign languages to domestic and international employability. In order to achieve these objectives, foreign languages are taught in Austria from the first year of primary school until the end of compulsory schooling as a minimum. As in much of Europe, English is the dominant 'first foreign language' learned (BMUKK, 2012, Teil II, Anlage 1, our translation). Recent statistics reveal that 91.8% of students at Austrian lower secondary level learn at least one foreign language and 8.1% two languages. Nearly all of those students who study foreign languages at lower secondary level learn English (99.9%)either as the 'first' or 'second' foreign language, while 1-5% learn French, Spanish, and Italian and less than 1% learn national minority languages like Slovene (Eurostat, 2019). While English is portrayed as the first foreign language in policy documents, as it is the first foreign language officially introduced at school, for students with German as an additional language, English may be a third or fourth language. Much concern has been expressed about whether the introduction of English is overwhelming for students who are still learning German, and research has suggested that teachers are sceptical of this simultaneous language learning (De Angelis, 2011, p. 223). However, blocking access to English is also deemed as inequitable, as it is often assumed that English education can be a stepping-stone to access better paid work and higher education (Solga & Dombrowski, 2009, p. 15), and English clearly has a privileged position within socio-politically established language hierarchies (De Swaan, 2001;Erling & Seargeant, 2013).
In Austria, students from lower socioeconomic status are less likely to have access to high quality school programmes which support English language education, as can be found in higher prestige schools (e.g. bilingual education, CLIL, preparation courses for international exams). They may also have less access to extracurricular, home resources than students from higher socioeconomic status (e.g. private language institutes, study abroad, travel, international visitors, English language resources) (Smit & Schwarz, 2019). Moreover, studies in other English-learning contexts have shown that the value of English in terms of enhancing socioeconomic status is related to other variables and systemic inequities in society in general (e.g. ethnicity, gender, language background) (Erling, 2017;Erling et al., 2019). Educational structures such as the dual-track system in Austria also seem to relate to educational outcomes in English language learning (Erling et al., 2020).

Educational tracking
A substantial obstacle to achieving equity in Austria is the early tracking system, where students are separated along the lines of their grades during their transition from elementary to lower secondary school. International research has established that inequity in education is significantly higher in tracked school systems than in comprehensive schooling systems (Hanushek & Wößmann, 2006). Austria is one of the countries in Europe with the shortest span of comprehensive schooling. As mentioned above, after four years of primary school, students are tracked into one of two different school types: middle schools, which aim to equip students for starting apprenticeships and/or moving on to more vocationally oriented upper secondary schools, and general academic secondary schools, which prepare students for university (BMBWF, 2018). While there is a certain degree of mobility and flexibility within the Austrian system, with students theoretically being able to move to any form of higher secondary education, only one out of ten students transfer from middle school to the academic track of higher secondary school, with many of those who make the transition dropping out in the first year (Beer, 2019). The majority of middle school students either only complete compulsory schooling or continue with vocational education (Statistik Austria, 2019). Thus, tracking a student into middle school is therefore often decisive for a student's entire academic career (Schwarz et al., 2002).
The rationale for tracking is to provide a homogeneous learning environment and an appropriate curriculum which enables maximum learning for all students in the classroom (Hanushek & Wößmann, 2006, p. 63). However, studies of dual-track systems in countries such as Austria and Germany have shown that the early tracking decision is strongly determined by the students' socioeconomic status, ethnicity and language background rather than their cognitive abilities (Leitgöb et al., 2015;Maaz et al., 2008;Oberwimmer et al., 2019). The assignment of students to the secondary track is not conducted by standardised norms but rather by the teachers' and parents' perception of the students' abilities, therefore giving a clear advantage to students whose parents are more familiar with the educational system (Maaz et al., 2008, p. 99). In fact, it has been found that hardly any other industrial state's educational performance depends so much on the parents' wages and educational background as it does in Austria (Becker, 2004;Maaz et al., 2008;Solga & Dombrowski, 2009). Early tracking systems as can be found in Austria have been found to exacerbate inequity, contribute to an increasing gap in educational attainment and create social differences between schools and their students (Baysu & de Valk, 2012;Hanushek & Wößmann, 2006;Maaz et al., 2008). While there are many middle schools of excellent quality that furnish students with a wealth of opportunities, students with low socioeconomic status, from a migration background and/or with German as an additional language are overrepresented in the non-academic track of schooling. This results in them having fewer opportunities for integration and social advancement. As a result, they may have less optimistic beliefs about their own learning and potential as well as lower aspirations (Lee, 2014).

Plurilingual pedagogies and equity
Previous research has also shown that teachers' beliefs about plurilingual practices are often negative and can therefore impede the implementation of equitable pedagogies drawing on these (see, for example, Haukås, 2016;Lundberg, 2019;Young, 2014). In this study, we therefore analyse the beliefs and classroom practices of English teachers at general academic secondary schools, and how they relate to students' learning outcomes. We then compare them to the beliefs and practices of middle school teachers. In doing so, we focus on the following research questions: (1) What are teachers' beliefs about their students' abilities to learn English, their students' motivations to learn English and their students' access to English outside of school? (2) What are teachers' reported language practices in the classroom?
(3) Do the beliefs and practices of general academic secondary school English teachers differ from those of middle school English teachers? If so, how?

Methods
The data analysed in this study was collected through a questionnaire that was circulated to teachers of English at general academic secondary schools in Austria from March-May 2020. Throughout this article, findings from this questionnaire are compared to those from a similar published study with teachers of English at middle schools conducted from March-May 2019 (for details, see Erling et al., 2020). Based on national laws and university statutes and guidelines, no ethics approval was required for the current study. The study adhered to the principles stated in the Declaration of Helsinki and complied with the British Association for Applied Linguistics Recommendations for Good Practice in Applied Linguistics, which are intended to help applied linguists to maintain high standards.

Participants
The two sets of participants were rather similar: Fifty academic secondary school teachers of English participated in this study (88% female; average age = 39 years, age range = 25-63) and fifty-six middle school teachers of English (93% female, with more than half being under 34). Forty-three teachers (86%) responded that they are German L1 users, while another four teachers are English L1 users, and three teachers provided other languages as their L1 (Albanian, French and Russian). Thus, the group of academic secondary school teachers was slightly more linguistically diverse than the group of middle school teachers, who all reported being German L1 users. Most of the academic secondary school teachers (86%) work in schools in the federal state of Styria, 10% in Lower Austria, and only a few in Upper Austria (2%) and Carinthia (2%). This compares to the middle school teachers, most of whom also work in schools in Styria (77%), with 11% in Vienna, and only a few in Upper Austria (4%), Lower Austria (5%), Carinthia (2%) and Burgenland (2%). Twenty-four of the academic secondary school teachers (48%) teach in an urban area (which is likely to be more diverse and multilingual), seven (14%) in a sub-urban area and nineteen (38%) in a rural area. For the middle school teachers, 45% teach in an urban area, 18% in a suburban area and 38% in a rural area, thus showing a similar distribution.

Materials and procedure
Teachers responded to a 92-item questionnaire that was almost identical to the one used in Erling et al. (2020). Questions were organised thematically and mostly required 5-point Likert-scale responses. The questionnaire covered demographic information, teachers' language practices in the classroom and teachers' beliefs about their students' language backgrounds and abilities, motivations for learning English, and access to English outside of school. Participants were recruited through various professional networks via email and social media. They followed a link to the questionnaire, which collected responses anonymously to protect teachers' identities.

Analysis
Responses were analysed using descriptive and inferential statistics. For the inferential statistics, we ran multiple linear regression analyses and t-tests. For all analyses, we assumed that the distance between the ordinal categories of our Likert-scale data is roughly equal. Such an assumption is reasonable for 5-point Likert scales with a neutral midpoint, and allows us to convert the ordinal data into numerical data for analysis. The linear regression analyses were the same as in the analysis of the data from middle school teachers (Erling et al., 2020). All linear regression analyses had the proportion of students that teachers believe to be achieving the learning outcomes for English (true for all/ most/about half/some/none of my students) as the dependent variable. The independent variables for each analysis were centred before analysis to minimise collinearity (Belsley et al., 2005). Independent variables that did not contribute significantly to model fit were removed in a stepwise procedure (Baayen, 2008). We report the results of these systematic model comparisons, i.e. results of the final analysis models, throughout. The p-values derived from the t-tests were corrected for multiple comparisons using a false discovery correction (Benjamini & Hochberg, 1995). The complete questionnaire, anonymized data, and analysis scripts for the current study are available on the Open Science Framework at https://osf.io/n4te7/.

Results
In the following sections, we explore how teachers' beliefs about their students' language background, abilities, motivations to learn English, and access to English outside of school as well as how teachers' reported language practices in the classroom relate to general academic secondary school students achieving the learning outcomes in English. We also compare these results to the analogous analysis for middle schools (Erling et al., 2020).
Teachers' beliefs about their students' language background by location Overall, the questionnaire revealed that 90% of the teachers (n = 45) indicated that all or most of their students are achieving the English learning outcomes for their level of schooling. This percentage is higher than the results of the 2019 standardised national assessment for English, which found that 69% of academic secondary school students are achieving the target learning outcomes of CEFR-level B1 or above in English in all of the areas of competence (BIFIE, 2020, p. 75). It is also much higher than the 30% of middle school teachers who indicated this. A t-test showed that perceived student achievement in the current study is significantly higher than amongst the middle school teachers (t = −7.61, df = 95.62, p = 0.002). The actual learning outcomes are indeed higher at academic secondary schools (69% vs. 27% of students meeting them in all areas of competence; BIFIE, 2020). However, the actual difference according to the national assessment is not as great as the perceived difference in the current study. Table 1 shows the percentage of students that teachers in rural, sub-urban and urban areas estimate as having German as an additional language. A large majority of teachers (80%) estimated that 10% or less of their students have German as an additional language, suggesting either that overall there is very little (perceived) linguistic diversity in the general academic secondary school classroom or that teachers are not fully aware of their students' language repertoires (cf. Brummer, 2019). The teachers reported a broad variety of languages spoken by their students, the most common of which were Bosnian/ Croatian/Serbian (BCS), Turkish, Arabic, Russian, Kurdish and Slovene. This roughly aligns with national statistics showing that on average 14% of academic secondary students have German as an additional language (BIFIE, 2020). This compares to a substantially higher number of middle school students with German as an additional language (24%).
We performed a multiple regression analysis with the percentage of students with German as an additional language at the school (10% or less, roughly 25%, roughly 50%, roughly 75%, 90% or more) and the location of the school (urban, sub-urban, rural) as independent variables to explore how these factors relate to the percentage of students that teachers believed to be achieving the learning outcomes for English. The results showed that the percentage of students with German as an additional language at the school and the school location significantly affect perceived learning outcome when considered together (β = 0.18, SE = 0.07, t = 2.34, p = 0.024). In contrast, the percentage of students with German as an additional language (β = −0.04, SE = 0.07, t = −0.50, p = 0.620) or school location (β = 0.08, SE = 0.07, t = 1.10, p = 0.276) individually did not significantly affect perceived learning outcomes. To explore how the percentage of students with German as an additional language and school location combine to affect learning outcomes, we conducted separate analyses for urban vs. rural and sub-urban schools. For urban schools, the final model contained no independent variables, such that we found no evidence that the percentage of students with German as an additional language at the school affected perceived learning outcomes. In contrast, in rural and sub-urban schools there was a significant main effect of the percentage of students with German as an additional language on learning outcomes (β = −0.21, SE = 0.07, t = −3.02, p = 0.006), suggesting that the higher the percentage of students with German as an additional language at these schools, the fewer students are believed to be achieving the learning outcomes. In the results for middle school teachers, the higher the percentage of students with German as an additional language related to fewer students perceived to be achieving the learning outcomes, irrespective of school location (Erling et al., 2020).
Teachers' beliefs about their students' language learning outcomes and abilities Overall, teachers had very positive views about their students' abilities to learn languages, with all teachers believing that at least half of their students were good language learners, and 76% of teachers even stating that most of their students were good at learning languages. This contrasts starkly with the results from middle school teachers, none of whom reported that all of their students were good language learners. A t-test showed that perceived language learning ability in the current study is significantly higher than perceived language learning ability at middle schools (t = −8.58, df = 92.98, p < 0.001). A vast majority of the academic secondary school teachers believe that English language education supports their students in developing a better sense of language, with 92% (n = 46) agreeing with this statement and 58% (n = 29) even strongly agreeing. These teachers were less convinced that learning English supports learning German for students with German as an additional language: Only about a third of the respondents (32%, n = 16) agreed with this statement, with almost half (44%, n = 22) being uncertain, and about a quarter (24%, n = 12) disagreeing. T-tests showed that agreement with these statements did not differ for teachers at academic secondary schools compared to middle school teachers (t = −0.53, df = 98.59, p = 0.596 and t = 1.56, df = 90.22, p = 0.135, respectively).
A multiple regression analysis further investigated whether the above factors influence learning outcomes. The independent variables for the analysis were teachers' beliefs about (a) whether learning English helps their students to develop a better sense of language in general, (b) whether learning English supports learning German for their students with German as an additional language (both: strongly agree, agree, neither agree nor disagree, disagree, strongly disagree) and (c) whether their students are good language learners (true for none/some/about half/most/all of my students). The final model contained no independent variables, suggesting that none of these factors influenced perceived learning outcomes in English. Overall, academic secondary school teachers' beliefs about their students' language learning abilities do not seem to impact students' learning outcomes, possibly because teachers had positive views of their students' abilities across the board. This contrasts with the results for middle school, where the more teachers believed that learning English helps their students develop a better sense of language and that their students are good language learners, the more they also believed that students were achieving the learning outcomes for English (cf. Erling et al., 2020). This finding might therefore suggest that the more teachers believe that plurilingualism is a resource and helps students develop a better sense of language, the more likely it is to be used as a resource that supports students achieving their learning outcomes.

Teachers' beliefs about their students' levels of motivation
Teachers overwhelmingly indicate that their students like learning English, with 90% (n = 45) agreeing that most or all of their students like English and 84% (n = 42) agreeing that most or all of their students are motivated to learn English. Academic secondary school teachers show significantly higher agreement with both statements when compared to middle school teachers, suggesting that they believe that their students are more motivated to learn English than middle school teachers do (t = −3.44, df = 80.86, p < 0.001 and t = −4.85, df = 93.65, p < 0.001, respectively).
We conducted a multiple regression analysis investigating whether teachers' beliefs about students liking English and being motivated to learn English relate to perceived learning outcomes (both: true for all/most/about half/some/none of my students). We found a significant main effect of motivation to learn English (β = 0.35, SE = 0.16, t = 2.24, p = 0.030): The more teachers believe their students to be motivated to learn English, the more they also believe that students were achieving the learning outcomes for English, either because students' higher motivation leads to higher learning outcomes or because lower learning outcomes in English lead teachers to believe that their students are less motivated. In contrast, there was no evidence that teachers' beliefs about how much their students liked English affected perceived learning outcomes. These results mirror those found in Erling et al. (2020) for middle school teachers.

Teachers' beliefs about students' access to English outside of school
Here, we consider teachers' beliefs about students' access to English outside of school as a possible contributor to the disparity in English language learning outcomes. Teachers overall had positive beliefs about their students' use of English outside the classroom: 22% (n = 11) indicated that most or all of their students regularly use English outside the classroom, with another 30% (n = 15) of teachers stating that about half of their students do so. Only one teacher (2%) believed that none of their students regularly uses English outside the classroom. This differs from the middle school teachers, where only 7% of teachers thought that most or all of their students regularly use English outside the classroom and 16% stated that none of their students do.
An overwhelming 90% (n = 45) of academic secondary school teachers believed that at least some of their students engaged in hobbies and extracurricular activities in English, compared to 71% of middle school teachers. Similarly, 100% (n = 50) of teachers stated that at least some of their students regularly use English for travelling and holidays, compared to 84% of middle school teachers. Overall, academic secondary school teachers show significantly higher agreement with all three statements regarding the use of English outside of the classroom as compared to middle school teachers (t = −3.96, df = 97.38, p < 0.001, t = −3.05, df = 101.19, p = 0.004, and t = −6.94, df = 99.39, p < 0.001, respectively).
A multiple regression analysis additionally explored whether teachers' agreement with the statements that their students (a) regularly use English outside of the classroom, (b) engage in hobbies and extracurricular activities in which English is used, and (c) regularly use English for travelling or holidays (all: true for all/most/about half/ some/none of my students) relates to perceived learning outcomes. The results reveal that all three independent variables affected learning outcomes in some way. Specifically, higher agreement with the statement that their students used English for travelling and holidays significantly related to higher reported learning outcomes for English (β = 0.19, SE = 0.08, t = 2.35, p = 0.023). No such independent effects on learning outcomes were found for using English outside of the classroom (β = −0.17, SE = 0.10, t = −1.75, p = 0.088) and for engaging in hobbies and extracurricular activities in which English is used (β = 0.04, SE = 0.10, t = 0.42, p = 0.674). Instead, these two variables need to be considered together with the variable of using English for travelling and holidays. If students were perceived to be using English outside the classroom less, then travelling was perceived to be more beneficial for their English learning (β = 0.28, SE = 0.10, t = 2.71, p = 0.010). If students were perceived to be more engaged in hobbies that use English, then travelling was perceived to be more beneficial (β = −0.31, SE = 0.11, t = −2.89, p = 0.006). The relationships among these variables can best be seen when splitting teachers into separate groups, based on their responses: A larger number of students using English for travelling and holidays positively affected learning outcomes only for teachers who said that half or fewer of their students used English outside of the classroom (β = 0.30, SE = 0.12, t = 2.46, p = 0.022), but not for teachers who stated that most or all of their students used English outside of the classroom (β = 0.09, SE = 0.11, t = 0.78, p = 0.441). This suggests that the positive effect on learning outcomes of using English for travelling is more evident if students do not usually use English outside of the classroom. Additionally, a larger number of students using English for travelling and holidays positively affected learning outcomes only for teachers who said that most or all of their students engaged in hobbies and extracurricular activities in English (β = 0.29, SE = 0.12, t = 2.49, p = 0.022), but not for teachers who stated that half or fewer of their students engaged in hobbies and extracurricular activities in English (β = 0.11, SE = 0.11, t = 1.03, p = 0.314). It is not immediately evident why this would be the case, so further studies to explore the relationship between travel and hobbies and English learning outcomes are needed. These rather complex results for academic secondary school teachers differ slightly from the results for middle school teachers, which showed independent positive effects of using English for travelling and holidays as well as using English outside of the classroom on reported learning outcomes.
Teachers' reported language practices in the classroom Our analysis of reported language practices in the classroom focuses on the use of German and other L1s in the English language classroom. 58% of academic secondary school teachers (n = 29) disagreed or strongly disagreed that they sometimes use German in the classroom, with another 20% (n = 10) neither agreeing nor disagreeing and 22% (n = 11) agreeing or strongly agreeing that they sometimes use German in the classroom. Overall, we find no evidence that academic secondary school teachers show significantly lower agreement with this statement (t = 1.94, df = 99.77, p = 0.069) compared to middle school teachers, though the comparison just failed to reach significance.
We found that 62% (n = 32) of teachers disagreed or strongly disagreed that they sometimes do activities in the English classroom in which students can use German, with only 16% (n = 8) agreeing with this statement. Furthermore, 34% (n = 17) of teachers found the statement that they sometimes do activities in the English classroom in which students with German as an additional language can use any of their languages not relevant. 45% (n = 15) of the remaining 33 teachers disagreed or strongly disagreed with this statement, 27% (n = 9) neither agreed nor disagreed, and another 27% (n = 9) agreed or strongly agreed. These findings cannot be compared to middle school teachers, as the statements were added to the questionnaire for the current study.
A multiple linear regression analysis further explored whether teachers' agreement with the statements that they (a) use German in the classroom, (b) sometimes do activities in which students can use German, and (c) sometimes do activities in which students with German as an additional language can use any of their languages (all three: strongly agree/agree/neither agree nor disagree/disagree/strongly disagree) relates to learning outcomes. We found no evidence that agreement with the above three statements influenced learning outcomes. This contrasts with the results for middle school teachers, where higher reported use of German in the classroom related to fewer students reportedly achieving the learning outcomes for English.

Discussion
In Erling et al. (2020), we argued that factors such as teachers' language use in the classroom, their beliefs about their students' abilities, their students' motivations and their access to English outside of the classroom might all be contributing factors to the lower outcomes in English at Austrian middle schools. In the current study, we start to see how these teachers' beliefs vary along the lines of the dual-track education system and how this variance might be contributing to inequity in English language education.
Although there were similarities, we found significant differences in our comparison of academic secondary school and middle school teachers' beliefs and classroom practices. The first major difference was that a vast majority of academic secondary school teachers indicated that all or most of their students are achieving the English learning outcomes for their level of schooling, compared to only 30% of middle school teachers. While national assessments show that there is a difference in the learning outcomes along the two school types, it is not as great as the perceived difference found in this study. Academic secondary school teachers also had very positive views of their students' abilities to learn languages, while their counterparts in middle schools did not. Such findings suggest that the difference in learning outcomes between students in the two school types may be exacerbated by teachers' differing beliefs about their students' language competences.
Another noteworthy difference was that in middle schools, it was found across the board that the higher the percentage of students with German as an additional language in a class, the more likely it was that teachers perceived fewer students to be achieving the learning outcomes. This perceived negative effect of having higher numbers of students with German as an additional language on learning outcomes was also noted by academic secondary school teachersalbeit only those in rural and sub-urban schools (and not those in urban schools). Thus, amongst academic secondary school teachers, negative beliefs about multilingualism might be more prevalent in rural and sub-urban areas, while they are equally common amongst middle school teachers in all areas. Academic secondary school teachers also had more positive views of their students' language learning abilities as compared to middle school teachers, and they are more likely to believe that learning English helps their students develop a better sense of language. Such positive beliefs that language learning and plurilingualism is a resource seem to have a relationship with students achieving the learning outcomes for English, also for the middle school teachers.
Academic secondary school teachers show significantly higher agreement with the statements that their students like learning English and that most or all of their students are motivated to learn English when compared to middle school teachers. In both studies, the more teachers believe that their students are motivated to learn English, the more they also believe that students are achieving the learning outcomes for English. While a relationship can be determined, causality cannot: either students' higher motivation leads to higher learning outcomes or lower learning outcomes in English lead teachers to believe that their students are less motivated.
Academic secondary school teachers overall had more positive beliefs about their students use of English outside the classroom. A much higher percentage of academic secondary school teachers believed that their students engaged in hobbies and extracurricular activities in English and regularly used English for travelling and holidays when compared to middle school teachers. In both studies, the more teachers agreed with the statement that their students used English for travelling and holidays, the higher students' reported learning outcomes for English were. In academic secondary schools, the positive effect of using English for travelling on learning outcomes is more evident if students are not perceived to be using English outside of the classroom. This indicates that teachers perceive travel as a means of making up for regular contact with English outside of school. This contrasts with the results for the middle school teachers, who perceived their students as having much less contact with English outside of school and who did not perceive travel as a means of making up for this.
Responses from teachers in both school types suggested that the majority do sometimes use German in the classroom, althoughin both studiesthere were quite a number who were unsure about this. Amongst the academic secondary school teachers, only a small number reported that they sometimes do activities in the English classroom in which students can use German. Almost half disagree that they sometimes do activities in the English classroom in which students with German as an additional language can use any of their languages. A third of them found this statement irrelevant. There was no evidence that agreement with the above three statements influenced learning outcomes. This contrasts with the results for middle school teachers, where higher reported use of German in the classroom related to fewer students reportedly achieving the learning outcomes for English. Teachers in both school types were not convinced that learning English supports learning German for students with German as an additional language. These findings suggest that in both school types, there is limited awareness and use of plurilingual pedagogies in which students can use their entire language repertoire to develop their English. While the middle school teachers acknowledged using German in English classes to a greater extent than their academic secondary school counterparts, the use of German did not seem to support their students achieving their learning outcomes. Perhaps if students' full plurilingual repertoires were mobilised (i.e. also their other home languages), this would better support English learning than the use of German only. Plurilingual pedagogies did not seem to be commonly used by the academic secondary schools either, so this is unlikely to explain the perceived difference in learning outcomes between the two school types. Other factors seem to be intervening here. However, given that teachers had far less confidence in middle school students' abilities to learn English, valuing and engaging these students' full language repertoires through plurilingual pedagogies might serve to enhance their confidence and the development of positive self-concepts, potentially allowing them opportunities to achieve higher outcomes that are more in line with their academic secondary school peers.

Conclusion
While further research is clearly needed in this area, the results of this study provide enough evidence to suggest a need for transformation in beliefs about multilingualism and the promotion of plurilingual approaches that would better contribute to socially just language education. By using students' multilingualism as a resource, enhancing intercultural communication, supporting the development of language awareness and cross linguistic transfer, teachers in the English language classroom could support students in the development of more positive self-concepts about their abilities as language learners, to help them see reasons to learn English, to identify and value their uses of English, thereby positively influencing these students' relationship to school-based (language) learning and enhancing their learning outcomes in English.
Enhanced outcomes in English learning in middle schools would be welcome and would go some way in redressing inequities in the education system. However, previous research has recognised how the value of English can be restricted by other variables such as ethnicity, gender and language background (Erling, 2017;Erling et al., 2019). English language education, compared to other school subjects, may have a particular stake and potential in drawing upon students' multilingualism and cultural diversity for the benefit of academic attainment (Moore, 2018). As national standards assessments show, large proportions of middle school students who are already multilingual and have intercultural experience are performing below their peers at academic secondary schools in English language outcomes (BIFIE, 2020). This may also be related to students' extracurricular and free-time activities, which have been found to play a central role in school achievement in Austria (Bacher, 2005, p. 57;Leitgöb et al., 2015, p. 37). The role of extracurricular activities in English language learning outcomes is also suggested in our study, which shows that students who have regular contact with English outside the classroom, through travel and hobbies, are more likely to be achieving learning outcomes. Moreover, opportunities to travelwhich are more likely to be out of reach for students from lower socioeconomic statusare perceived by teachers as being able to make up for regular access to English outside of the classroom.
The findings of this study are in line with the argument that improving outcomes in English is important, but that alone is unlikely to enhance equity in Austria, as other socioeconomic and systemic factors play important roles in constructing educational disadvantage. The Austrian tracked school system is an inter-and intragenerational reproduction of inequity, which cannot be counteracted by individual policy changes or educational reforms (Solga & Dombrowski, 2009, p. 7). In Oberwimmer et al. (2019, p. 132), the Ministry of Education openly recognises that Austria is one of very few countries which still have a dual-track system and that in many cases, the students' educational career is not determined by their performance but rather by their social and ethnic origins. Inequity caused by socioeconomic backgrounds is not being compensated for at school, but rather perpetuated, and being magnified by policies, practices and beliefs about the tracked educational systems. The variance in beliefs and practices in English language learning across school types seems only to be contributing to gaps in education, but plurilingual pedagogies might present an opportunity to support transformations in the direction of more equitable education in English language education and beyond. Since plurilingual pedagogies are not yet commonly employed in either school type, their potential for transformation in English language education requires further implementation and investigation. Note 1. The variable of socioeconomic status is an indexical measure derived from parents' educational and professional status and the reported number of educational resources like books at home (BIFIE, 2020).