Keywords

1 Introduction

The main goal of foreign language teaching is to prepare learners to use the language in formal and informal settings of social interaction in order to co-construct meaning (Council of Europe, 2001; Larsen-Freeman & Anderson, 2011; Thornbury, 2011). The foreign language lesson can be viewed as a social setting in which teacher and learners engage in interaction around a certain topic, for instance derived from a text. A meta-analysis (Murphy et al., 2009) revealed that active student engagement in classroom discussions about a text promotes co-construction of meaning. However, these authors also state that the way in which classroom discussions are organized matters greatly. An important prerequisite for effective discussions is that the teacher does not dominate the discussion, but that there is room for students to express thoughts, ideas and feelings during classroom interaction (Murphy et al., 2009). According to Murphy et al. (2009), it is not so much the quantity but the quality of classroom discussions that matters greatly in achieving co-construction of meaning.

Many researchers have acknowledged the importance of fostering co-construction of meaning in the language classroom (Gibbons, 2015; Walqui & Van Lier, 2010; Mercer & Dörnyei, 2020). However, classroom dynamics may be influenced by a host of factors, for instance student ability, number of students in the classroom, lesson topic and type of classroom activities (Dewaele, 2020; Mercer & Dörnyei, 2020; Dörnyei et al., 2015). These factors might impact the extent to which co-construction of meaning between teacher and students is achieved. A recent observational study (Smit et al., 2022) focused on what teachers and learners do to foster co-construction of meaning during interaction and revealed a gap between what is happening in classrooms and what research says about effective classroom interaction. The study provided systematic descriptions of teacher and learner question and answer behavior, and operationalized co-construction of meaning as active participation in question and answer sequences by everyone in the classroom most of the time. Asking questions is one of the basic tools in a teacher’s pedagogical repertoire (Murphy et al., 2009). A teacher’s open-ended question (i.e. no predetermined answer) can serve as an invitation for learners to contribute to co-construction of meaning. Smit et al.  (2022) found highly active teachers and rather inactive students.

An important question with regard to educational research and teaching practice is to what extent they might inform each other. Research findings are not always understood, recognized or deemed relevant by practitioners. The general aim of this study was to bridge the theory-practice gap. The observational study of Smit et al. (2022) did not reveal underlying factors for the observed behavior. The first aim of the present study was to find out whether teachers who were not observed but work in the same teaching context in The Netherlands, think the observational evidence is representative of actual practice. The second aim was to investigate how teachers in The Netherlands would attribute the observations and what they thought might improve teacher-student interaction patterns in EFL lessons in the Netherlands.

2 Literature Review

2.1 English as a Foreign Language Teaching in the Netherlands

English is one of the three core curriculum subjects in the Dutch curriculum for secondary education. Communicative foreign language teaching forms the backbone of the national curriculum for English as a foreign language (EFL) (Fasoglio et al., 2015). The Dutch curriculum has been aligned with the Common European Framework of Reference (hence CEFR) and requires from 15–18-year-old students that they are able to enter discussions about a wide range of both familiar and unfamiliar topics at CEFR level B1+ / B2 (Fasoglio et al., 2015; Council of Europe, 2001). Understanding texts also plays a major role in the Dutch curriculum. By the end of secondary education, Dutch learners in the highest levelsFootnote 1 take a national standardized reading exam at CEFR level C1 (Fasoglio et al., 2015). This exam determines 50% of the final grade for English. These curricular requirements illustrate why it is important for Dutch teenagers to be able to read English texts and discuss these texts during foreign language lessons at school.

2.2 From Observations to Perceptions of Classroom Interaction: The Role of Lesson Content, Teaching Materials and Language Proficiency

Factors that have been suggested as a major influence on how the language lesson unfolds are lesson content (i.e. what is talked about), teaching materials, and learners’ language proficiency (Thornbury, 2011; Larsen-Freeman & Anderson, 2011). Regarding the content of the language lesson, the discussion is complicated. In a language lesson any topic could be approached from a language learning perspective, but according to Arnold (1999) it is crucial that the subject matter is appealing and relevant to the language learners. In order to foster learner engagement, proposals have been made to incorporate learner-oriented topics in the lessons (Maley, 2011). However, when this was operationalized as using lesson content derived from popular culture (e.g. film, music, celebrities) to focus on grammatical structures, this did not automatically lead to increased learner engagement (Piggott, 2019; Lightbown, 2015; Dönszelmann et al., 2020).

Considering the second factor, teaching materials, recent studies have shown that the coursebook determines what happens in Dutch EFL lessons (Tammenga-Helmantel & Maijala, 2019) and that a heavy focus on restricted language practice does not help learners to interact in real-time (Van Batenburg et al., 2018). Additionally, these studies revealed large amounts of cognitively and sometimes also linguistically unchallenging discourse (Van Batenburg et al., 2018; Tammenga-Helmantel & Maijala, 2019). This suggests a possible gap between the way coursebooks prepare students for social interaction and the skills that are needed for actual social interaction inside and outside the classroom.

Thirdly, in order to interact with other people in another language, sufficient lexico-grammatical knowledge as well as a sufficient level of oral fluency are needed (Council of Europe, 2018). A study of oral fluency levels of Dutch teenagers by Fasoglio and Tuin (2017) confirmed that students in the two highest levels of Dutch secondary education attain the desired proficiency level, i.e. CEFR B1-B2 for speaking. Moreover, this study showed that a large proportion (48.6%) of the students in pre-university education achieve CEFR C1 level for oral fluency. An important additional finding from this study was that Dutch teenagers, although fluent enough, often do not use the English language in the classroom. In a sample of teenagers in pre-university education (n = 385), 20% of the students reported never to attempt to only use English as the language of communication during classroom interaction. These results suggest that active classroom participation is not a precondition for students to achieve relatively high fluency levels. Only 10% of the students in pre-university education always try and use English during the language lesson. In lower levels of secondary education, the percentage of students who speak English in class was even lower (Fasoglio & Tuin, 2017). Although Dutch teenagers seem to be reasonably fluent in English, they show limited evidence using the language in the classroom.

2.3 From Observations to Perceptions of Classroom Interaction: The Role of Emotions and Motivation in Classroom Interaction

The fourth and fifth factor that might impact classroom interaction relate to affective aspects in the language learning process (Arnold, 1999). We will discuss both emotions and motivation in relation to Self-Determination Theory, hence SDT (Deci & Ryan, 2000). SDT focuses on what moves people into action by describing human psychological needs in terms of relatedness, autonomy and competence. Gibbons (2015) illustrates competence and relatedness by discussing the role of emotions and stresses that a certain amount of struggle in understanding others and making yourself understood is needed to get ahead in language learning. She also points out that moments of frustration are most significant when learners are communicating with “a helpful interactant” (Gibbons, 2015). However, when frustration causes students to lose confidence and feel embarrassed or anxious, learning stalls. According to Dewaele et al. (2018) lessons which are emotionally uninteresting or emotion-free, might lead to routine, boredom and lack of engagement, which could suggest a weak sense of relatedness.

A student who is bored might try to avoid active participation, but a lack of response from the learners could in turn influence the teacher’s sense of relatedness and competence, which in turn could affect interaction. Although proficiency levels of qualified English teachers in the Netherlands are at CEFR C1/C2 (10 Voor de Leraar, 2018) and there is no evidence that teacher proficiency might be a limiting factor, Dönszelmann (2019) reports that foreign language teachers confessed to struggle being consistent in their use of the foreign language during the lessons. Whereas linguistic competence might not be at stake, a threat to relatedness or experienced autonomy and teaching competence might play a role here. Underlying emotional factors for this struggle to use the English language consistently might be that teachers’ worry that students do not understand what they are saying, or that students and parents might complain about the intelligibility of the language lesson (Fasoglio & Tuin, 2017; Dönszelmann, 2019).

Finally, learner motivation might also impact classroom interaction. Language learning motivation might fluctuate during the lesson and these fluctuations could impact the quality and quantity of student participation during the language lesson (Waninge et al., 2014). Research into language learning motivation has focused on factors such as the value and relevance for the language user, being able to use the language, and the goals learners want to achieve (e.g. educational or professional advantages) (Dörnyei et al., 2015). These factors also relate to SDT’s relatedness and competence (Deci & Ryan, 2000), constructs which are closely associated with cognitive, emotional and behavioral engagement (Mercer & Dörnyei, 2020). National surveys revealed that Dutch teenagers have a positive attitude towards the English language and its relevance (Fasoglio & Tuin, 2017). Based on Dutch teenagers’ self-reported levels of emotional engagement with the English language would suggest sufficient motivation to learn this language. However the multi-dimensional and dynamic nature of this construct (Waninge et al., 2014) might also implicate that sufficient motivation might not directly lead to active verbal student behavior during classroom interaction.

2.4 Observed EFL Classroom Interaction and Teacher Cognition

Teacher cognition research seeks to address the relationship between what teachers do in their teaching practice and what they think, know and believe. This type of research is often carried out to complement classroom observational research (Borg, 2006; Basturkmen, 2012). Johnson (2006) stresses that teacher cognitions and pedagogical decisions mutually influence each other and change over time. It is therefore important to examine both teaching behavior, which is defined here as what teachers do during their lessons, and teachers’ perceptions of the observed behavior.

Questions and answers are building blocks of social interaction that can be observed and labelled relatively clearly and were therefore chosen by Smit et al. (2022) as a representation of moment-to-moment teacher-student interaction patterns that occur naturally in a language lesson. The results from this observational study revealed that teacher questions and student answers have the tendency to form patterns dominated by closed teacher questions and simple student answers. During a 50-minute lesson, English as a foreign language teachers asked around 60 questions on average to which students gave short (i.e. one to three-word utterances) or no answers. Micro-level observations also revealed that in 30% of the lessons (n = 16), students had the tendency to adjust the level of their answer to the level of the teacher question (e.g. ‘low level’ questions leading to ‘low level’ answers, higher level questions leading to higher level answers). However, this study found no evidence for a relation between the teachers’ follow-up question and the previous student answer. The study provided detailed descriptions of the micro-dynamics of teacher-student interaction in foreign language lessons, but did not yield insight in underlying reasons for the observed interaction patterns (Smit et al., 2022).

3 The Present Study

The first aim of the present study was to find out whether teachers think the observational evidence found in Smit et al. (2022) is representative of actual teaching practice. The second aim was to investigate how teachers would attribute the observed patterns and what they would suggest as directions to improve teacher-student interaction patterns in EFL lessons in the Netherlands. The present study was designed to minimize attribution errors that might be caused by the actor-observer effect of confirmation bias. Teachers may have varying reasons for choosing to participate in an observational research study. However, the presence of a camera in the classroom might influence teacher and student behavior, making it difficult to determine to what extent the observations are “business as usual”. Therefore teachers from the same teaching context who had not been observed were asked to participate in this study. The study seeks to answer the following research questions:

  1. 1.

    Do teachers recognize the observed interaction patterns that are characterized by a dominance of closed teacher questions and short student answers in their own teaching practice?

  2. 2.

    What is the best explanation for the observed classroom interaction according to EFL teachers?

  3. 3.

    What do teachers perceive to be the best suggestion for improving teacher-student classroom interaction?

4 Method

4.1 Participants and Context

Teachers (n = 57) attending a presentation about classroom interaction were asked to participate in a short questionnaire about the classroom observational evidence. The data was presented and explained by the first author of this paper on two different occasions in January and March 2020. The first group of respondents (n = 47) were EFL teachers participating in a teacher conference organized by the University of Groningen in January 2020. One of the conference participants was not a teacher, but worked as a consultant for an educational publisher. This respondent was excluded from the study. The second group of respondents (n = 10) were trainee teachers in the Master of Education at the University of Groningen attending a seminar about interaction in the language classroom. This seminar was part of an English language teaching methodology course taught by the first author of this paper. During their masters’ program the trainee teachers also worked as EFL teachers in schools for secondary education in the Netherlands.

All respondents in this study (n = 57) were familiar with the EFL teaching context in Dutch secondary education and had hands-on teaching experience. Respondents were asked to answer our questions as if it were their own practice. The response rate for completing the anonymous questionnaire was 100%, which might be due to the convenience sampling procedure described above and the short amount of time needed to complete the questionnaire (less than 3 min on average). All participants were first asked for consent to participate and were given the possibility to opt out immediately. The research design was approved by the Ethics Committee of the Department of Teacher Education at the University of Groningen (EC reference 19-024/RM/AA).

The sample consisted of respondents working in different levels of Dutch education. A large majority (86%) of the respondents was female, 12% were male and one person (2%) indicated “other” for gender. An overall majority of the respondents were EFL teachers working with teenagers in two highest levels of Dutch secondary educationFootnote 2 (43 people – 74%), 13 teachers (24%) taught English in (pre)vocational secondary education (teenage learners), one person (2%) worked as an EFL teacher in higher education (young adult learners, >17 years old) . The distribution between experienced and early career professionals (defined as anyone who had between 0 and 5 years of experience) was roughly two-thirds (35 people – 61%) to one third (22 people – 39%). This means that the majority of the respondents who reflected on the classroom observational evidence that was presented during the presentation had substantial experience teaching learners of a similar age and educational level (i.e. higher secondary and pre-university education).

4.2 Procedure

At the start of the presentation, the first author of this paper explained the relevance of classroom interaction research and provided some background information about the context of the research project. The teachers were informed that observational data in Dutch secondary education classrooms had been collected in lessons taught to learners (14–17 years old) preparing for higher vocational or university education. All observed lessons used a text as a language input, which meant that lessons with a focus on teaching grammar were excluded from this study. It was explained that classroom interaction had been studied by observing sequences of teacher questions and students’ answers and that teacher questions and student answers had been coded with the Questions and Answers in English Language Teaching (QAELT) coding scheme (Smit et al., 2022). This coding scheme consists of four-point scales for teacher questions and student answers in which openness and level of complexity are accounted for. Table 34.1 displays the simplified version of QAELT coding scheme as presented to the respondents.

Table 34.1 Simplified version of QAELT coding scheme
Full size table

After explaining the coding system, the observational evidence was presented. For the representation of the observational data three State Space Grid visualizations (Hollenstein, 2013; Lamey et al., 2004) were used. The scale for teacher questions is displayed on the horizontal axis of the State Space Grid and the vertical axis displays the scale for student answers. Together these scales form a 4x4 grid. Every dot in the grid represents an interaction which is formed by a teacher question combined with a student answer. The respondents were first informed that the “closed question – simple answer” pattern was the dominant pattern for the majority of the observed lessons (5 out of 16 lessons, i.e. 31%). The closed question-simple answer cell is the region in which most interactions took place. Then a State Space Grid showing a lesson with high levels of interaction and a different type of dominant pattern was presented to the respondents. This was the state space grid of lesson d4 displayed in Fig. 34.1. The grid of lesson d4 reveals that the teacher received an answer to every question. Additionally, the majority of the questions in this lesson took place at the level of clarification or open-ended questions.

Fig. 34.1
A scatterplot of student answers versus questions on a 4 cross 4 chess pattern. The Y-axis headers are complex, complete, simple, and none. The X-axis headers are non-elicitation, closed, clarification, and open-ended. The maximum dots are plotted at the (clarification, complete) grid.

State Space Grid visualization of a lesson with high levels of interaction (lesson d4)

Full size image

Next, the teachers looked at a lesson (a1) with a low level of interaction (see Fig. 34.2). In this lesson the closed question and the simple answer, indicated by the yellow box, was the dominant pattern. Notably, a lot of questions that were asked during this lesson did not receive an answer at all.

Fig. 34.2
A scatterplot of student answers versus questions on a 4 cross 4 chess pattern. The Y-axis headers are complex, complete, simple, and none. The X-axis headers are non-elicitation, closed, clarification, and open-ended. The maximum dots are plotted at the (closed, simple) grid.

State Space Grid visualization of a lesson with low levels of interaction (lesson a1)

Full size image

Finally, teachers gauged State Space Grid b2 (Fig. 34.3) which depicted the median level of observed interaction in EFL lessons from the data set that was used in Smit et al. (2022). In order to establish the median level of teacher-student interaction the following measures were used: number of questions and percentage of questions in most the frequently occurring cell of the State Space Grid. The median number of teacher questions uttered during a 50-minute lesson in the dataset was 51. The most frequently occurring cell in this data set was the closed question – simple answer cell. The lesson with the median percentage of interactions (26%) in this cell was lesson b2. From a sample of 16 lessons, seven lessons had a lower percentage of interactions in the dominant cell and eight lessons had a higher percentage in the dominant cell. It was explained to the respondents that we chose to show the median level of observed interaction in order to validate the sample median. We asked the respondents whether they thought the level of interaction in their lessons was either lower or the same, or higher than the median level of interaction in the sample. It was explained to the respondents that lesson b2 represented a lesson “in the middle”, represented by the median.

Fig. 34.3
A scatterplot of student answers versus questions on a 4 cross 4 chess pattern. The Y-axis headers are complex, complete, simple, and none. The X-axis headers are non-elicitation, closed, clarification, and open-ended. The maximum dots are plotted at the (closed, simple) grid.

State Space Grid visualisation of a lesson with the median level of interaction (lesson b2)

Full size image

The respondents filled out the digital anonymous Qualtrics (hhtps://www.qualtrics.com) questionnaire immediately after the presentation. The questionnaire could be accessed by the participants by using a QR code or a shortened url. After filling out consent, gender, teaching experience and type of school in which the teachers worked, they were asked to answer the questions in Table 34.2 based on their expertise.

Table 34.2 Questionnaire about teachers’ perceptions
Full size table

The questionnaire was designed in such a way that there was a relation between the answer options to question 3 and the answer options to question 5. Question 3 consisted of possible explanations for the observation classroom observation patterns and question 5 consisted of possible measures for improvement aligned with the explanations. Table 34.3 shows how the answer options of these two questions correspond.

Table 34.3 Explanations for classroom interaction and possible measures to improve
Full size table

From Table 34.3 it can be seen that dedicating classroom attention to vocabulary and conversation skills was suggested in order to address possible language learning issues. Making students feel more competent (for instance by using formative evaluation techniques) was proposed to overcome possible emotional barriers. Problems in lesson content might be addressed by teaching about topics that students are interested in. A solution for teaching materials that do not encourage learners enough to participate actively would be to make teaching materials more interesting. And finally, motivational factors, for instance students who do not want to learn English at school, could be targeted by actively increasing students’ motivation to learn English. Both questions 3 and 5 had a forced response, which means that participants were asked to pick only one explanation and only one measure.

Immediately after filling in the questionnaire, group results for all questions were displayed to the respondents, after which the first author of this paper and the respondents engaged in a brief discussion about the results. The goal of this discussion was teacher development and therefore not included in this study.

5 Results

Regarding the research question (RQ1) whether teachers recognize the dominant patterns of classroom interaction, an overall majority of the respondents (82%) confirmed that the observations were in line with their expectations. A small minority (7%) indicated that the results were worse than they had expected, and 11% indicated that this was better than they had expected. When the teachers were asked if they thought classroom could be improved, almost all respondents (96%) said ‘yes’ and only two (4%) believed that improvement was not possible. Regarding the teachers’ self-assessment of interaction patterns in their own lessons the results show that 72% of the respondents thought that the level of classroom interaction in their lessons is similar or lower to the observed median level of interaction. The results show that roughly a third (30%) of the respondents indicated that the level of interaction in their lessons is higher.

With regard to the question of what the best explanation for the most frequently observed patterns of classroom interaction was (RQ2), a majority (72%) attributed the observed interactions patterns to emotional factors (see Fig. 34.4). According to 14% of the respondents, a lack of encouraging teaching materials is the best explanation for the observed results. This means that most respondents suggested that emotional factors play an important role in the emergence of classroom interaction patterns that are characterized by active teachers asking many closed questions and inactive students giving no answers or very short answers.

Fig. 34.4
A horizontal line graph plots 5 bars for language skills, emotional factors, lesson content, teaching materials, and motivational factors. Their respective values are 3.51%, 71.93%, 5.26%, 14.04%, and 5.26%.

Best explanation for classroom interaction according to the participants

Full size image

Further analyses of the responses revealed that a large majority (81%) of the experienced (>5 years) teachers attributed the observed interaction patterns to emotional factors. A smaller majority (59%) of the inexperienced teachers (0–5 years) thought that emotional factors were the best explanation for the observed patterns. One in three (31%) inexperienced teachers mentioned that the content and teaching materials could be a possible explanation for relatively inactive learners.

Regarding the possibility for improvement (RQ3), 98% thought improvement was possible. The results of the follow-up question (Table 34.2, question 5) about measures to improve classroom interaction are displayed in Fig. 34.5. The proposed measures to improve classroom interaction were increasing attention for vocabulary and conversation skills, making students feel more competent (formative evaluation), teaching about topics that interest the students, making teaching materials more interesting and increasing the motivation for learning English. Making students feel more competent by using formative evaluation was the most promising measure according to the respondents (51%). Making teaching materials more interesting was also suggested (18%), one respondent (2%) thought that classroom interaction could not be improved, incorporating more interesting topics were suggested by four teachers (7%), six teachers proposed increasing motivation (10%) and seven teachers (12%) preferred the option to improve vocabulary and conversation skills.

Fig. 34.5
A horizontal line graph plots 6 bars for not possible, vocabulary and conversation skills, teaching materials, interesting topics, formative evaluation, and increase motivation. Their respective values are 1.75%, 12.28%, 17.54%, 7.02%, 50.88%, and 10.53%.

Measures to improve classroom interaction

Full size image

In the final question of the questionnaire teachers were also offered the opportunity to indicate how they thought classroom interaction could be improved. Nineteen respondents (33%) answered this question. The suggestions provided by the respondents could be linked to the following five broad themes: classroom organization (47%), the national curriculum with a focus on starting early (11%), lesson content (5%), professional development (11%) and teaching materials (26%). Table 34.4 gives an overview of the themes, the number of comments made and for every theme one illustration of the answers given by the respondents.

Table 34.4 Qualitative analysis of answers to the open question
Full size table

Suggestions regarding improvements in classroom organization, especially the importance of a safe classroom climate were given most often as an additional solution for the lack of student activity. Teaching materials were mentioned by the respondents who opted for emotional factors in the closed question and who also indicated that more factors might play a role. Teaching materials were also mentioned in relation to using technology and digital tools.

6 Discussion and Conclusion

A group of EFL teachers who had not previously been observed were asked to reflect on observational findings on classroom interaction in their teaching context in The Netherlands. A very large majority of the respondents (82%) recognized the observed patterns, which could indicate that interaction patterns characterized by active teachers and inactive students might be a familiar struggle for many teachers in the Netherlands. The respondents were presented with observations of a lesson with a median level of interaction and we asked them whether their lessons had higher levels of interaction or the same or lower. Overall, respondents indicated that the observed interaction patterns confirmed their expectations of classroom dynamics regarding teacher questions and student answers. Only a third of the respondents thought that the level of active student participation during classroom interaction in their lessons was higher, which could imply that average levels of active student participation in EFL lessons in the Netherlands might be somewhat lower than observed. A large majority of the respondents believed that classroom interaction can be improved.

From the literature, we know that joint attention and joint action are important mechanisms to achieve co-construction of meaning in the language classroom (Allwright, 1984; Larsen-Freeman & Cameron, 2008). Our respondents attributed the lack of students’ responsiveness to teacher questions mainly to emotional factors. Some of the respondents suggested that a lack of student responsiveness might be due to classroom routines which are not conducive to language development. An example of such classroom routines are situations in which a language teacher asks questions which students can easily answer, or moments during the lesson in which teachers accept short answers. These asymmetric interaction patterns can be frustrating for teachers and potentially boring or uncomfortable for teenagers. According to Gibbons (2015) frequent interactions characterized by closed teacher questions and simple student answers could be characterized as a “high support/low challenge” interaction.

Whether teachers and learners actually are conscious of their own behavior (i.e. closed questions, simple answers) in real-time and the potential effect this might have on lessons, we do not know. A possible explanation might be that it is cognitively too demanding for teachers to monitor both a large group of students and themselves during the teaching-learning process. However, suggestions provided by the respondents indicate that teachers who might consciously or unconsciously work hard to maintain a safe learning climate, could also lead to routines in which teachers avoid putting teenagers on the spot by pushing for more extensive verbal output in English.

Learners who let their teacher to do most of the talking might implicitly shift the responsibility for managing the interaction to the teacher. From the perspective of teenage students, this might be an attractive option: limiting the amount of what you say can be an effective way to reduce risk of entering a potentially awkward, difficult or embarrassing situation in which you lose face in front of your peers. The benefits for teenagers of merely showing the teacher that they are “on board” by just listening and giving short but correct answers are high. This suggests that in whole class teacher-student interaction both learners and teacher could benefit from adhering to a relatively traditional distribution of authority. Future research, for instance observations of interpersonal behavior (Pennings et al., 2014) combined with a stimulated-recall interview, might look into whether the implicit agreement, the teacher leads and talks, whereas the students follow and answer, exists.

In order to overcome potentially uncomfortable situations, the respondents in this study offered some practical solutions such as asking questions but also using digital tools to let all students first give an anonymous online answer, before entering a classroom discussion. The respondents argued that this might lower the threshold for students. Adopting classroom management techniques to maximize active participation might offer suggestions to improve the balance between levels of teacher and student activity (Scrivener, 2012; Mercer & Dörnyei, 2020).

Research in the field of positive psychology suggests that fostering positive emotions can enhance language learning (MacIntyre et al., 2019; Dewaele, 2020). However, ignoring negative emotions like frustration, embarrassment and boredom and failing to address these might result in suboptimal behavioral patterns that are hard to change. Acknowledging that negative emotions are part of the learning process and offering opportunities to fail and learn from frustration might be needed to pave the way for positive experiences of learning and development and fostering relatedness (Gibbons, 2015; Deci & Ryan, 2000). In order to overcome suboptimal patterns of teacher-student interaction, a small majority of the respondents proposed to invest in formative evaluation practices. Formative evaluation is focused on getting ahead by providing ongoing interactive feedback during the process of learning. Process feedback might simultaneously address the basic needs of relatedness, autonomy and competence: hel** students understand their current level of communicative competence, offering suggestions to change real-time behavior in order to become more autonomous, whilst hel** each other in getting ahead by kee** the classroom conversation going.

It is promising that teachers recognize emotional struggles and suggest that researchers direct their attention to the cognitive and affective domain of learning simultaneously (The Douglas Fir Group, 2016). It is also promising that teachers express a wish to better understand and change classroom interaction. This study has shown that asking teachers to reflect on observational evidence of interaction patterns might improve their understanding classroom interaction and encourage them to reconsider how to make the most of the teacher’s turn.