Teachers’ self-reported L1 and L2 use and self-assessed L2 proficiency in primary EFL education

This study investigates teachers’ first language (L1, German) and second language (L2, English) use in the primary English as a foreign language (EFL) classroom in two federal German states. It particularly focuses on the question of whether a more frequent, (self-reported) use of the L2 is positively correlated to teachers’ professional qualification as well as (self-assessed) L2 proficiency. To this end, data was collected in 2017 through an online survey among German primary teachers teaching EFL in year 4 (N = 844). L2 use was assessed through a 4-point Likert scale comprising 16 items on various classroom situations. L1 use was surveyed with an open question on situations of L1 use in the L2 classroom. Moreover, teachers self-assessed their L2 proficiency with a 4-point Likert scale and adapted CEFR descriptors for speaking. Findings indicate that teachers claim to use the L2 more in L2-related situations and the L1 more in classroom management situations. The study shows that teachers with a higher formal qualification tend to assess their L2 proficiency higher and claim to use the L2 more often in the primary EFL classroom. In contrast, teachers with a lower formal qualification tend to assess their L2 proficiency lower and claim to use the L1 more frequently in the L2 classroom.


Introduction
The guiding principle "as much L2 as possible, as little L1 as necessary" has been widely accepted for instructed foreign language learning (e.g., Butzkamm & Caldwell, 2009). Proponents argue that native language (L1) use can support learning of the foreign language (L2). In contrast, there is the danger of teachers overusing the L1 and thus disadvantaging their learners by providing less L2 input and fewer learning opportunities (Turnbull, 2001). Prior studies found negative correlations between frequent L1 use and learners' L2 proficiency (Helmke et al., 2008). Thus, L2 teachers should make a conscious decision in which situations and for what purposes they revert to the L1. English as a foreign language (EFL) teachers' use of the L1 in the L2 classroom can be explained by factors such as their L2 proficiency or beliefs (e.g., McMillan & Rivers, 2011). Moreover, teachers' perceptions of their learners' needs, including learners' L2 proficiency, impact their uses of the L1 or L2. Studies investigating out-of-field L2 teachers, that is, teachers who did not major in an L2, found that they use the L2 less frequently, have a lower L2 proficiency (Dörr, 2018), and feel insecure when using the L2 (Deters-Philipp, 2018). The present study was conducted in the German context of primary EFL education. It aims to explore in which situations primary EFL teachers report using the L1 (German) and the L2 (English). It also investigates if more frequent use of the L2 is positively correlated to the teachers' professional qualifications as well as their self-assessed L2 proficiency.

Contextual and theoretical background
The context of primary EFL education and teacher education in Germany will first be sketched out to illustrate the background of the study. Second, findings from prior studies regarding primary EFL teacher education as well as L1 and L2 use will be summarized.

Primary EFL education in Germany
In Germany, the federal states (Länder) are responsible for the provision of education. Consequently, the 16 German federal states have their own school system with their different curricula as well as teacher training and development agendas. However, in all of the 16 federal states children enter primary education at the age of six and, with the exception of the federal states of Berlin and Brandenburg, transfer to secondary education at the end of year 4.
Since the school year 2004/2005, foreign language education -in most states EFL -has been obligatory in primary schools across Germany. As of 2020, children begin learning a foreign language in the majority of federal states in year 3 and in four states as early as year 1 (Rixon, 2013). Data for the present study was collected online from primary EFL teachers in two federal states with different ages of onset: whereas in Lower Saxony (LoS) primary school children begin learning EFL in year 3, at approximately 8 years of age, they start as early as year 1 in North-Rhine Westphalia (NRW), when learners are approximately 6 years old. In both states, curricula stipulate oral skills as one of the main objectives of early EFL education (MSW NRW, 2008;NKM, 2006). Regarding written skills, however, the curricula are slightly different. Whereas the LoS curriculum emphasizes the oral skills (NKM, 2006), the NRW curriculum prescribes a more pronounced integration of written language: Teachers are required to provide written language input to support EFL learning right from the start. Furthermore, in contrast to the LoS curriculum, the NRW curriculum specifies the expected EFL competence levels for both oral and written skills at the end of primary education in year 4 (MSW NRW, 2008).
The overarching goal of the political decision to introduce primary foreign language education was to further consolidate learners' language proficiency and, thus, meet the goals of European language policy. However, both the introduction of primary foreign language education as well as the introduction of primary foreign language teacher education has not been empirically evaluated on a large scale (e.g., Porsch & Wilden, 2017).

Primary EFL teacher education in Germany
Teacher education in Germany is structured in two phases (Cortina & Thames, 2013). In the initial phase, prospective teachers study for a master's degree (or its equivalent). In the second phase, they complete a school-based pre-service training. For primary teacher education, the 16 federal states follow different policies concerning the school subjects future primary teachers are required to study. In almost all states student teachers study three subjects. Also, in most states there are requirements as to the subjects they can choose to study (Porsch, 2017). In most states, such as for example NRW, trainee primary teachers are obliged to study German (as the language of schooling) and mathematics in addition to a third primary subject. As a result, the route to becoming a primary teacher is very diverse across Germany.
With the introduction of compulsory primary foreign language education across Germany, this diversity of formal teacher qualifications has been further increased by the introduction of yet another option. At the time that this modification came into effect there was an immense lack of qualified foreign language teachers. Thus, the federal governments provided short-term, intensive post-qualification courses to qualify teachers with other majors for teaching primary EFL. It should be assumed that to this day the majority of EFL learners in German primary schools will learn the language from a teacher who did not major in EFL education. In the context of primary EFL education it is empirically unclear which effects different formal teacher qualifications have on their learners' achievement.
Internationally, a number of studies have investigated whether teacher certification matters (Goldhaber & Brewer, 2000). These studies are principally based on the assumption that subject-related competence acquired by teachers during their initial teacher education is essential for effective teaching (cf. Garton, 2019). This leads to the assumption that learners who learn from teachers qualified and certified for a subject -also called in-field teachers or specialized teachers -are advantaged over those learners taught by out-of-field teachers, or teachers who are formally qualified for a subject or "field" other than the one they are teaching.
In the context of primary education in Germany, teaching primary EFL outof-field happens regularly due to two reasons: a dramatic lack of teachers trained as specialists (see above) and the so-called class teacher principle (Porsch, 2020). The latter means that in a given class the same primary teacher will teach almost all subjects. In this context, it should be noted that in most universities student teachers electing to study primary EFL will be expected to have L2 proficiency at the C1 level. Teachers without an EFL certificate can opt to attend an intensive post-qualification course in order to obtain a subject-specific teaching certificate. However, these courses mainly focus on teaching methodology and do not assess or develop teachers' L2 proficiency. In sum, primary EFL teachers with at least three different types of qualifications currently work in German primary schools: 1) those who majored in EFL, 2) teachers who majored in other subjects and participated in a post-qualification course for teaching EFL, and 3) teachers who majored in other subjects and did not undergo any training in teaching EFL. These teachers' L2 proficiency is not tested at any point; however, primary teachers who majored in EFL are likely to be at the C1 level or higher.
Several studies have identified subject-specific teacher qualification as a factor associated with improved learning outcomes of pupils in this particular subject (e.g., Clotfelter, Ladd, & Vigdor, 2010). However, other studies yielded contradictory evidence and were not able to explain learner achievement by indicators of teachers' qualification (e.g., Porsch & Wendt, 2017). The inconclusive nature of prior findings possibly originates in different research designs (see Porsch & Whanell, 2019) and, more importantly, varying definitions of what constitutes an "out-of-field" teacher or who counts as a "qualified/non-qualified teacher." The cohort of out-of-field teachers is rather diverse with some "pursuing an interest" (Hobbs, 2013, p. 294) by teaching a new subject and possessing subject-specific knowledge, whereas others are "just filling in" (Hobbs, 2013, p. 294). Except for individual characteristics, the provision of support by colleagues and school management potentially reduces the challenges for outof-field teachers (see Hobbs, 2013).
In EFL education, proficiency in the target language is an essential element of the professional competence of teachers. Thus, in the context of EFL education, learners' achievement depends largely on their teachers' L2 skills. In fact, studies indicate the significance of teachers' L2 proficiency for the development of their pupils' language skills (e.g., Unsworth, Persson, Prins, & de Bot, 2015). In an interview study in German primary schools, Deters-Philipp (2018) found that EFL teachers' confidence in speaking English is closely related to their formal qualifications. Out-of-field teachers tend to feel less secure when using the L2. Another study in lower secondary EFL education using data from classroom observations and teacher interviews found that out-of-field-teachers use the L2 less and have a lower L2 proficiency than those with a major in EFL (Dörr, 2018). This is in line with a longitudinal observation study in Mexico by Izquierdo, García Martínez, Garza Pulido, and Aquino Zúñiga (2016), which found that secondary teachers' L2 proficiency seemed to play a key role in their classroom L2 use. As these findings suggest, formal teacher qualification ought to be considered a factor in early EFL education and the same should apply to its effect on actual classroom teaching and children's learning outcomes.

L1 and L2 use in (primary) EFL education
In this section, prior findings regarding L1 and L2 use in the EFL classroom will be reviewed. First, the focus will be on studies investigating actual classroom practice. Second, studies investigating learners' and teachers' perspective on language use will be summarized. The review focuses particularly on studies in the context of EFL education for young learners. However, in some instances the authors also included ESL contexts or older learners as they directly relate to the research focus of the present study.
Across Germany, curricula for primary EFL education stipulate the L2 as the predominant means of communication in the foreign language classroom (KMK, 2013). However, they suggest functional language use, which means that the L1 may be used as well to support target language learning. This is in line with findings from second language acquisition studies as well as learning psychology which show that the L1 is a central element in acquiring a new language. Large-scale observations studies in Germany found that secondary EFL teachers switch to the L1 on a regular basis and in different situations (DESI-Konsortium, 2008;Ott, 2011). Even though the frequency of L1 use by the teachers varies, most teachers claim to be insecure about the potential benefits or disadvantages of their own L1 use for their pupils (Ott, 2011). Similarly, the longitudinal observation study by Izquierdo et al. (2016) conducted in 37 Mexican secondary schools provided evidence that English language teaching is characterized by overreliance on the L1. In many contexts of instructed foreign language learning, teachers are the main source of target language input and function as main role models. Thus, too much L1 use by the teacher is likely to reduce L2 learning opportunities and, hence, might have a negative impact on the development of their target language proficiency.
On the other hand, there is ample evidence supporting the notion that L1 use in the L2 classroom can support L2 learning. For example, previous studies found that teachers successfully use the L1 to explain the meaning of new words or a grammatical structure (e.g., Deters-Philipp, 2018;Hall & Cook, 2013;La Campa & Nassaji, 2009;Macaro, Tian, & Chu, 2018;Nakatsukasa & Loewen, 2015;Paker & Karaağaç, 2015;Sali, 2014). In doing so, they use the L1 as a scaffold (Bhooth, Azman, & Ismail, 2014) and to exploit the L1-L2 connections learners often make in their minds (McMillan & Rivers, 2011). It was found that code-switching can support learning target vocabulary and can in particular benefit intermediate learners (Lee & Levine, 2018; also see Tian & Macaro, 2012). Furthermore, an observation and interview study by La Campa and Nassaji (2009) found that teachers sometimes use the L1 to facilitate classroom communication and thus support the development of learners' L2 communicative competence (also see Atkinson, 1987;Auerbach, 1993).
Prior studies focusing on the learners' perspective have provided evidence that learners favor the inclusion of the L1 in the L2 classroom. For example, Macaro and Lee (2013) in a survey (plus follow-up interviews with a subsample) of elementary and adult learners of English in South Korea found that both younger and older learners favor the inclusion of the L1 in the classroom (also see . Particularly relevant in this context, the study found that younger learners welcomed L1 use much more than the adult learners. Similarly, the findings by Lee and Lo (2017) point to the necessity of carefully considering the particular target group when determining the amount of L1 and L2 use. The study found that learners' ideal L2 self (i.e., the vision of themselves as L2 users in the future) was a stronger predictor of their attitudes towards classroom language choice than their L2 proficiency.
Shifting the focus to teachers' perspectives on language choice in the L2 classroom, it is interesting to consider the survey among Japanese university teachers of English by McMillan and Rivers (2011), who found that, contrary to official policy, many teachers believed that occasional use of their learners' L1 could enhance their L2 learning. Similarly, La Campa and Nassaji (2009) offered evidence that teachers believe in the benefit of using the L1 to facilitate L2 learning.
In this context, it is interesting to focus on teachers' reasons for L1 use. For example, Deters-Philipp (2018) found that teachers sometimes feel forced to switch to the L1, for example, on account of reasons of classroom organization, discipline issues or to avoid misunderstandings.
Based on the available evidence, Cummins (2007, p. 227) summarizes that there is "no empirical justification for any absolute exclusion of learners' L1 from TL [i.e., target language] instruction." Even in monolingual classrooms, "[like] nature, the L1 creeps back in, however many times you throw it out with a pitchfork" (Cook, 2001, p. 405). Thus, in Germany as well as in many other contexts the following guiding principle has been widely accepted (e.g., Bhooth et al., 2014;Cameron, 2001;Rolin-Ianziti & Varshney, 2008;Shabir, 2017): "as much L2 as possible, as little L1 as necessary." Also, there seems to be a general consensus that the amount of L1 use should slowly decrease while increasing classroom discourse in the L2 (Cameron, 2001;Hall & Cook, 2013). In this context, from a pedagogical point of view, Butzkamm and Caldwell (2009) argue that L1 use in L2 education allows teachers to create a safe and friendly learning atmosphere, in which learners gain confidence and become less dependent on their L1 (also see Auerbach, 1993). They characterize the L1 "as the most important ally a foreign language can have" (Butzkamm & Caldwell, 2009, p. 24), which values and recognizes the learners' L1 as a rich learning resource.
In view of the above considerations, the present study investigates primary EFL teachers' self-reported L1 and L2 use, with a special focus on teachers' self-assessed L2 proficiency and professional qualification. In doing so, the study distinguishes between L2-related situations and classroom management situations of L1 and L2 use in primary EFL education (Atkinson, 1987;Auerbach, 1993;Cameron, 2001;Cook, 2001;Hall & Cook, 2013;Rolin-Ianziti & Varshney, 2008). This is because the language used by the teacher in the classroom serves various functions. First, in foreign language education the L2 represents the content or subject-matter students learn. Second, the quality of language plays an important role for effective classroom management (Macías, 2018). Other functions are, for example, creating a relaxed learning atmosphere by, for example, using humor. L2-related use of the L1 refers to classroom interactions that focus on teaching or learning the target language. This could, for example, include the provision of corrective feedback in the L1 or situations of L1 use in which there is talk about characteristics of the L2 or meta-cognitive talk. This could also include using the L1 to support individual learners to keep up with the group. Classroom management situations of L1 use involve giving instructions, setting up activities or rearranging the classroom. The major objective of L1 use in this regard appears to be saving time as well as increasing time-on-task by ensuring that all learners can follow. In this context, Cameron (2001) encourages teachers to make a conscious decision about which language they use depending on what they consider supportive of learning outcomes. Similarly, Ott (2011) and Inbar-Lourie (2010) differentiate between motivated code-switching, that is, intended and conscious L1 use in L2 education, and performance switching, that is, spontaneous, unintentional and compensatory L1 use. These motives for teachers' L1 use in L2 education relate to very typical L2 classroom situations. In the present study, we distinguish between learner-related reasons and teacher-related reasons for L1 use in the L2 classroom. L2 teachers might either use the L1 because they respond to learners' needs and assume that L1 use in some situations supports individual learners better. In addition, L2 teachers might use the L1 from time to time in order to deal with managerial issues. This in turn may have to do with specific teacher characteristics and less with students' needs. For example, if teachers lack sufficient subject-specific training and do not have a high level of language proficiency, they potentially feel less confident in using the L2. Consequently, out-of-field teachers are more likely to use the L1 in stressful situations, for example, when they need to manage discipline issues.

Research questions
The present study examines the use of L1 and L2 in the context of EFL primary education in Germany. It extends earlier research by considering different professional qualifications of primary EFL teachers as well as their language proficiency in this regard. Teachers who are not qualified or not fully qualified to teach English or language teachers with a lower L2 language proficiency are more likely to switch to their L1 in the L2 classroom (Deters-Philipp, 2018;Dörr, 2018;Izquierdo et al., 2016). Therefore, the study addresses the following research questions: 1. In which classroom situations do primary EFL teachers report using the L2 (English) and in which situations the L1 (German)? 2. Which reasons do primary teachers give for their use of the L1 in the EFL classroom? 3. Are differences in the self-reported L1 or L2 use of primary EFL teachers related to different professional qualifications? 4. Is there a statistically significant relationship between the self-assessed language proficiency of primary EFL teachers and their self-reported L1 or L2 use in the classroom?

Participants and procedure
Data was collected in summer 2017 through a computer-based survey among German primary teachers teaching EFL in year 4 (N = 844). For all teachers that participated in the study, German was the L1 and English was the L2. All teachers responded to all closed items that were part of the questionnaire. In order to invite participants to the study, the authors e-mailed the school management of all primary schools in two federal states. In this e-mail the purpose of the study was explained and the school management was requested to forward the information on the study to EFL teachers at their school. Participation in the study was voluntary. Eight hundred and nine (N = 809) of the participating teachers were female (95.9 %). They were between 25 and 67 years old (M = 40.86, SD = 10.09). All of the teachers taught primary EFL at the time of the survey. With regard to their subject-specific qualification, the following groups can be distinguished: 339 teachers (40.1 %) had a major in EFL education, 404 participants (47.9 %) had majored in other subjects but participated in a post-qualification course in primary EFL, and 101 teachers (12 %) had majored in other subjects and obtained no further professional training to teach EFL.

Instruments
In order to answer the research questions, the following data from the questionnaire was considered for analysis: · L2 use in the EFL classroom: The 4-point Likert scale (1 = "never or almost never" to 4 = "always") comprised 13 items on various classroom situations (see Appendix A). Teachers were asked "How often do you use English in the following situations?" The items were adapted from two previous studies, the DESI-study and BIG-study, conducted in Germany (Barucki et al., 2015;Wagner, Helmke, & Rösner, 2009). However, neither of the studies considered a classification of situations. By considering findings from the qualitative analysis (see below), this study distinguishes between situations that focus on teaching the language (L2-related, 8 items, e.g., "to introduce new words," ɑ = .72) and those that refer to classroom management activities (classroom management, 5 items, e.g., "in case of disruptions," ɑ = .71). The correlation between the two scales was moderate (r = .58, p < .001), and thus we opted for a separability of the two dimensions instead of using one scale. · L1 use in the EFL classroom: The survey included the following openended question: "Are there any situations in which you speak German? And if so, in which?" 563 teachers answered the question. In total, 66 of the teachers answering this question had obtained no professional training for teaching EFL, 282 had completed a post-qualification course in primary EFL, and 215 had majored in EFL education. The authors can only speculate about why not all teachers responded to this question. It is possible that those who failed to do so were teachers who never used German in their L2 classroom. Others might not have answered because they considered L1 use as undesired and thus did not want to reveal their L1 use even if the survey was conducted anonymously. · EFL teachers' L2 proficiency: All teachers were asked to assess their proficiency in English (see Appendix A). To this end, CEFR descriptors for speaking were adapted (Council of Europe, 2001; levels A1 to C2; M = 3.38, SD = .56, 6 items, ɑ = .88). The teachers were asked the following question: "To what extent do these statements apply to your ability to speak English?" The teachers could answer on a 4-point Likert scale ranging from 1 ("strongly disagree") to 4 ("strongly agree").

Analysis
The statistical analyses were carried out using SPSS (version 25.0). To compare mean differences between two groups/scales we used t-tests with Bonferroni correction. A multivariate analysis of variance (MANOVA) was applied in order to estimate the effects of the independent variable ("qualification") on the dependent variables (the two situations). As the data from the open question was not normally distributed, non-parametric tests of variance (Kruskal-Wallis) were used to assess the statistical significance of group differences. The open answers regarding primary teachers' L1 use in the EFL classroom were analyzed with the help of structuring qualitative content analysis (QCA; Kohlbacher, 2006;Schreier, 2012) with the MAXQDA software package, which allows both an open and flexible as well as theory-guided and systematic approach to analyzing verbal data. The coding frame was developed in a deductiveinductive process right on this dataset (Zech, 2019). Consistency of the coding frame and data analysis was checked both through cyclical as well as consensual coding (Kohlbacher, 2006;Schreier, 2012). The final coding frame which was applied to the entire data set comprises six main categories which capture the different dimensions of L1 use in the EFL classroom (multiple coding possible): 1) L2-related situations, 2) classroom management situations, 3) other classroom situations, 4) unspecific or unclear answers, 5) learner related reasons, and 6) teacher-related reasons. Main categories 1 and 2 were developed theory-based regarding the L2-related and classroom management dimension of L1 use in the foreign language classroom. Main categories 3 and 4 were theory-guided as well as data-driven but will not be further considered here as they do not relate to the focus of the present study. Main categories 5 and 6 were derived from the data as participants' answers not only related to situations but in some instances also gave reasons for the their decision to use the L1 (see section 4.3). For each main category, a set of subcategories was developed in a deductive-inductive process in order to capture the different aspects of each dimension. As some responses were assigned to more than one category, there was a total of 1,122 codings. Sixty eight (N = 68) answers were unclear or unspecific and could not be further considered in the analysis (main category 4). One hundred and fifty seven (N = 157) answers were assigned to main category 3) other classroom situations.

Teachers' L2 use in the primary EFL classroom
The means for the two scales and all participants of the survey (see Table 1) indicated that teachers used the L2 (English) in their classroom most frequently in situations focusing on the target language and less often in classroom management situations. The difference was statistically significant (p < .001). The mean value of L2 usage in classroom management situations was comparably low (M < 3), but indicated that the teachers did use the L2 for classroom management purposes as well. A multivariate analysis of variance (MANOVA) showed a main effect for teacher qualification (F(2, 843) = 31.526, p = .001, ƞ 2 = .07 and F(2, 843) = 20.388, p = .001, ƞ 2 = .05), respectively. Comparisons between the three groups (t-tests with Bonferroni correction) confirmed that there were statistical differences of group mean scores (p < .001), with the exception of the means for L2 use between group 1 and group 2 in both L2-related (p = .086) and classroom management situations (p = 1.00). The mean comparisons suggested that a higher professional qualification in teaching EFL was related to a more frequent use of L2 in the primary school classroom.

Teachers' L1 use in the primary EFL classroom
The open answers regarding teachers' use of the L1 (German) in their EFL teaching were coded according to the QCA coding frame. In the following, the focus will be on findings in main categories (1) L2-related situations and (2) classroom management situations. In total, 351 teachers' answers were coded as L2-related (see Table 2) and 404 as classroom management situations (see Table 3) of L1 use in the primary EFL classroom. Thus, there was a slight majority of answers relating to aspects of classroom management. When comparing this dimension to L2-related situations, it is interesting to see that the latter were spread over several sub-categories: almost 30% of the teacher answers related to using the L1 to explain words, phrases or larger units of meaning, for example: "[I use German] when I don't get on with gestures and facial expressions. . ." (1.1; our translations). Approximately 26% of the answers related to explaining grammar aspects and about 12% referred to situations in which learners failed to mediate or translate to the L1, for example: "when there are huge difficulties in understanding and not even a very good pupil can translate" (1.4). About 15% of the answers referred to using the L1 to support individual learners in their L2 learning, for example: "after storytelling to ensure comprehension of weaker learners" (1.7). Fewer answers related to L1 use to talk about characteristics of the L2, about L2 learning processes or aspects of testing and assessment. In addition, the Kruskal-Wallis-test (a rank-based nonparametric test as an alternative to the one-way ANOVA) was used to test for statistically significant differences between the three groups in the frequency of L1 use in the various L2-related situations (codes 1.1 to 1.7). The analysis found no significant differences for L1 use in L2-related classroom situations between the groups of teachers with different professional qualifications (code 1.1: χ 2 (2) = 2.717, p = .257; code 1.2: χ 2 (2) = 3.939, p = .140; code 1.3: χ 2 (2) = 3.338, p = .188; 1.4: χ 2 (2) = 1.623, p = .444; 1.5: χ 2 (2) = 3.382, p = .184; 1.4: χ 2 (2) = 3.117, p = .210).
The classroom management dimension had two clear foci of L1 use in the primary EFL classroom. More than 48% of teacher answers related to aspects of giving instructions, such as explaining activities or setting up the classroom, for example: "for very complex tasks" (2.1). Almost 35% referred to situations linked to keeping up classroom discipline, for example: "when there are persistent disruptions of the lesson" (2.3). Again, the Kruskal-Wallis-test was applied to test for significant differences between the three groups of teachers with different professional qualifications regarding their L1 use in classroom management situations (codes 2.1 to 2.5). The analysis found significant differences between the three groups of teachers for the code 2.2 "Lesson transparency" (χ 2 (2) = 8.640, p = .013) and code 2.3 "Classroom discipline" (χ 2 (2) = 6.2425, p =.040). However, as there were no equal variances between the groups, pairwise multiple comparison were conducted by using Dunnett's C, a test that is appropriate when the variances are unequal. The analysis did not show any statistically significant differences. Also, for no other code were significant differences between the different groups identified (code 2.1: χ 2 (2) = 4.302, p = .116; code 2.4: χ 2 (2) = 1.572, p = .456; code 2.5: χ 2 (2) = 2.116, p = .347).

Reasons for teachers' L1 use in the primary EFL classroom
Analysis of teachers' answers regarding their L1 use in the primary EFL classroom showed that some also gave reasons for such use, even though the question had only asked for classroom situations. In the QCA, a total of 72 answers were assigned to main category learner-related reasons and 46 (see Table 4) to teacher-related reasons (see Table 5). In the learner-related dimension, the subcategories with most codings were support individual learners (n = 23, 31.9 %) and keep affective barriers low (n = 27, 37.5 %), for example: "to encourage them, reduce inhibitions" (5.5, our translations).  In the teacher-related dimension, the sub-category with most codings was saving time (N = 25; 54.3%), for example: "when I'm running out of time." In the context of this study, it was particularly interesting to note the small number of responses in which teachers quite openly admitted to using the L1 in order to compensate for a lack of L2 proficiency, for example: "when I don't know the right [English] word" (6.4; our translation). Because of the small number of responses regarding teachers' reasons for using the L1 in the L2 classroom, we do not report differences by teacher qualification in this regard.

Relationship between teachers' L2 proficiency and their L1 or L2 use
In the present study, teachers were asked to self-assess their L2 (English) proficiency for speaking in addition to reporting on their L1 and L2 use in the primary EFL classroom. The Pearson correlation between self-assessed L2 proficiency and self-reported target language use in L2-related situations was about .29 (p < .001) and in classroom management situations .32 (p < .001), which indicates a small positive relationship between these variables. In other words, the higher the teachers self-assess their L2 (English) proficiency, the more often they use the L2 in the primary EFL classroom.
In order to analyze the relationship between teachers' L2 proficiency and L1 use in the classroom, the quantitative and qualitative data from the present study were merged. The Spearman correlation shows only in one case a significant relationship between (the absolute frequency of) L1 (German) use in the EFL classroom and teachers' self-assessed L2 (English) proficiency significant on a 99% level: There was a negative correlation of -.15 (p < .001) between giving instructions as a classroom management situation and teachers' self-assessed L2 proficiency (all correlations are reported in the Appendix B). This indicates that the likelihood of giving instructions in the L1 (German) was higher the lower teachers self-assessed their L2 (English) proficiency for speaking.
Thus, to further explore the relationship between teachers' professional qualification for teaching EFL and their L2 proficiency, the correlation coefficient was estimated for these two variables. The analyses found a significant correlation of .39 (p < .001) between the two variables (professional qualification for EFL and L2 proficiency). This indicates that primary EFL teachers with a higher professional qualification for teaching EFL were those who self-assess their L2 proficiency higher; also, they tended to use the L1 less often in the EFL classroom than their colleagues who were formally less qualified.

Discussion
This study investigated teachers' use of L1 and L2 in the context of primary EFL education in two German federal states. It extends earlier research on language use in the L2 classroom by considering different professional qualifications of primary EFL teachers as well as their self-assessed language proficiency. To this end, primary EFL teachers were surveyed on their L1 (German) and L2 (English) use in the classroom, on their professional qualification and on their self-assessed L2 proficiency. Similar to previous studies (DESI-Konsortium, 2008;Izquierdo et al., 2016;Ott, 2011), the present investigation found that primary EFL teachers reported using both L1 and L2 in the L2 classroom. They reported to use the L2 more in classroom situations focusing on target language learning. In addition, the study found that primary EFL teachers claimed to use the L1 more in classroom management situations. These findings indicate that many primary EFL teachers make a conscious decision as to when they use the L1 or L2, as suggested by various authors (e.g., Cameron, 2001;Ott, 2011). The fact that the teachers claimed to use the L2 more in target language situations and the L1 more for classroom management indicates that teachers consider both the need to provide their learners with ample L2 learning opportunities and follow a learner-centered approach by using the L1 as a form of scaffolding. This is in line with findings from prior studies which identified teachers' positive attitudes towards L1 use in the L2 classroom (La Campa & Nassaji, 2009;McMillan & Rivers, 2011). At the same time, the findings are in line with previous findings that teachers sometimes feel forced to switch to the L1 for classroom management aspects (Deters-Philipp, 2018).
Focusing on primary EFL teachers' professional qualifications and selfassessed L2 proficiency, the study found the following: The higher primary EFL teachers' professional qualification, the higher they tend to self-assess their L2 proficiency and the more often they claim to use the L2 in the classroom. And, similarly, the lower primary EFL teachers' professional qualification, the lower they tend to self-assess their L2 proficiency and the more often they report to use the L1 in the classroom. These findings are in line with those in previous studies, which suggests that teachers with lower L2 proficiency use the L1 more often in the L2 classroom (Izquierdo et al., 2016). Furthermore, prior studies identified lower L2 proficiency among out-of-field (primary) EFL teachers, who feel less secure in using the L2 and use the target language less in their L2 teaching (Deters-Philipp, 2018;Dörr, 2018). Considering that teachers' target language proficiency seems to be significant for their pupils' L2 development (e.g., Unsworth et al., 2015), the apparent (self-assessed) lower L2 proficiency of outof-field EFL teachers should be a reason for concern among educational policymakers. In short, the less input students receive in the L2, the fewer learning opportunities they have, which can lower their potential learning growth. This assumption is supported by a large representative study from Germany which demonstrated that year 9 learners who were taught EFL by out-of-field teachers are considerably less proficient in L2 reading and listening comprehension than those taught by specialist EFL teachers (Hoffmann & Richter, 2016). This study has some shortcomings that need to be acknowledged. Thus, at the methodological level, the sample was not representative as participation was voluntary. In effect, the authors assume that primarily those teachers participated who were motivated and interested in the topic. Another limitation is the fact that, for pragmatic reasons, only self-reported data were used, particularly regarding the teachers' L2 proficiency but also regarding their actual L1 and L2 use. This relates to questions of validity, since teachers might have responded in a way they thought the researchers expected. Preferably, future research initiatives in this area ought to assess teachers' L2 proficiency (e.g., by using standardized language tests) and investigate their actual language use as well as teaching methods through classroom observations (e.g., by rating teachers' language use). At the content level, a number of limitations need to be considered as well. Firstly, the study only focused on teachers' use of the L1 and the L2. In doing so, it disregarded how teachers actually understand and integrate their learners' utterances in the L1. The L1 is an important variable in children's L2 learning processes. If EFL teachers are unable to comprehend their learners' L1, for example in the context of this study with children growing up with L1s other than German, this might disadvantage learners. Secondly, the teachers' open responses revealed a lot about their teaching methods and beliefs, but the exploration of that data was beyond the scope of this paper. As teachers' beliefs are closely linked to their teaching behavior, this aspect might be considered in future studies.

Conclusion
Findings of this study are highly relevant in the current context of primary EFL education in Germany, especially regarding the ongoing political and public debate on this school subject (Wilden & Porsch, 2020). For example, due to teacher shortage, the current federal government of NRW has specifically encouraged out-of-field teachers (so-called Seiteneinsteiger) to take up positions as primary EFL teachers. At the same time, it is considering plans to move the start of EFL education from year 1 back to year 3. In the light of this political debate, the authors suggest considering findings from the present study as arguments for the employment of fully qualified primary EFL teachers. Finally, further research initiatives are needed to investigate both teachers' L2 proficiency in relation to other aspects of primary EFL education and the effects of primary EFL teachers' qualifications on their learners' L2 proficiency. support in collecting and processing data for this study: Janina Ehmke (Essen), Natalie Hofstra (Essen), Benjamin Möbus (Vechta), Rebecca Schlieckmann (Paderborn), and Jule Zech (Essen). I can present clear, detailed descriptions of complex subjects integrating sub-themes, developing particular points and rounding off with an appropriate conclusion, I can present a clear, smoothly-flowing description or argument in a style appropriate to the context and with an effective logical structure which helps the recipient to notice and remember significant points.