Get it together
Smit, Nienke
DOI:
10.33612/diss.160498701
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2021
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Smit, N. (2021). Get it together: exploring the dynamics of teacher-student interaction in English as a
foreign language lessons. University of Groningen. https://doi.org/10.33612/diss.160498701
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Summary, discussion and general
conclusion: between the ideal and reality
6.1. Introduction
In a continuously changing multicultural and multilingual world, the English language is often used as a lingua franca in diverse socio-cultural professional and educational contexts (Council of Europe, 2018; The National Standards Collaborative Board, 2015). Preparing teenagers for a future in which they will able to use the English language meaningfully is therefore an important task for language teachers across the globe (Mercer & Dörnyei, 2020). In a seminal paper in The Modern Language Journal about second language acquisition in a multilingual world a group of scholars makes a plea for transdisciplinary studies that acknowledge the multi-faceted nature of language learning and teaching (Douglas Fir Group, 2016). The authors also encourage researchers to study the micro-level of social activity, because they argue that this is the level from which language learning and teaching starts.
This dissertation has adopted a holistic, transdisciplinary and complex dynamic systems perspective of learning and teaching English as a foreign language. Building on previous classroom observational studies in Dutch primary and secondary education (Van de Grift, 2007; Van de Pol et al., 2012; Menninga et al, 2017), the primary objective of this dissertation was to describe the dynamics of teacher-student interaction in English as a foreign language (EFL) lessons in Dutch secondary education. This dissertation has provided empirical descriptions gathered in naturalistic classrooms. The secondary objective of this study was to develop valid and reliable instruments that might be used by teacher educators and researchers to make real-time student-teacher interaction in the foreign language classroom measurable. In order to meet these aims two different observation procedures were developed to describe EFL classroom interaction. The way teacher-student interaction takes place during the lesson is a process which a teacher might try to organize by deliberately using a repertoire of teaching strategies that support student language development in reading comprehension lessons. This notion formed the starting point of the transdisciplinary study in chapter 2, in which insights
from educational effectiveness research were combined with applied linguistics research about foreign language teaching. From the results of Chapter 2, questions arose about the actual way EFL teacher-student interaction unfolds from moment to moment. Because this process can only be understood by observing the way teacher and learners respond to and influence each other in real-time on the micro-timescale of the lesson, Chapters 3, 4 and 5 are grounded in a complex dynamic systems research perspective.
6.2. Co-construction of meaning from a complex dynamic systems
perspective
The way language teachers and learners respond to each other during the lesson constructs and constrains language learning (Allwright, 1984). Ideally, the interactive process between teacher and learners during a lesson leads to learner language development and the teacher’s support is temporary, calibrated and adaptive. In order to develop their foreign language proficiency, students may benefit from communicating with the language teacher, who is often the most proficient speaker in the room (Gibbons, 2015). According to Walqui and Van Lier (2010) it is crucial for language development in a classroom setting to maximize verbal student participation. There are various ways in which teachers pursue the aim of “using language for co-construction of meaning” in language teaching. Teachers do many different things to support learners in many different ways, but one of the most common pedagogical tools for language teachers is to try and engage learners in elaborate verbal interaction during their lessons.
The interactive process between teachers and learners in which they use language to learn language to “construct” language development is often referred to as scaffolding (Van de Pol et al., 2010; Wood et al.,1976). The scaffolding metaphor is a popular concept for researchers and teachers, but due to its complex and dynamic nature it is very difficult to measure or quantify. Scaffolding is often defined as providing support at the right
moment in time, which points out the crucial importance for teachers to understand their learners and respond adequately (Van de Pol et al., 2010). Steenbeek and Van Geert (2005a) propose that teachers and learners are studied together whenever this kind of dynamic scaffolding behavior is the focus of a research study. These authors also argue that scaffolding is a dynamic transactional process unfolding in real-time and requires dense time-series data that capture moment-to-moment interactions.
In this dissertation, co-construction of meaning as a possible outcome of interactive teacher scaffolding was investigated by analyzing patterns emerging from questions and answers during EFL lessons (chapters 3 and 4). The process of co-adaptation between teacher and learners formed the focus of chapter 3, whereas the temporal dimension analyzed in terms of levels of stability and variability was the central theme in chapter 4. An important limitation of observational research in general, is that an exclusive focus is on teacher and learner behavior cannot uncover underlying perceptions, intentions or beliefs. Chapter 5 therefore focused on investigating teachers’ perceptions of the observed behaviors and interactions.
The coding procedures used in chapters 3 and 4 were inspired by research carried out in elementary education. Menninga (2017) studied question and answer patters in science lessons taught to young children. In this dissertation, moment-to-moment patterns emerging from teacher questions and student answers were quantified and analyzed using complex dynamic systems concepts such as self-organization, co- adaptation, attractor states, stability and variability. These concepts will be revisited in the summary of the four studies.
6.3. Teacher student interaction in the Dutch context
In order to understand what is happening in a real classroom adequate and accurate descriptions are key. Naturalistic classroom observations of ‘business as usual’ EFL lessons formed the backbone of this dissertation. In Dutch secondary education the duration of a foreign language lesson is limited to 50 minutes on average, two or three times a week. Additionally, Dutch classrooms are filled with around 25-30 teenage learners. Ultimately, teenage learners of English in the Netherlands should have sufficient communication skills (CEFR B2/C1) which they can use for co-construction of meaning in contexts of study, work or leisure. Although it should be acknowledged that there are various other relevant modes of classroom interaction, for instance small group work or pairs, the focus of this dissertation is limited to whole-class teacher student interaction because this is currently the most prevalent mode of interaction “business as usual EFL lessons” in Dutch secondary education. Because of this common practice, research into the process of teacher-student interaction aimed at supporting large groups of learners within the limited timeframe of the lesson is highly relevant. The teachers who participated in the observational studies were asked to teach an EFL lesson in which they use a fiction or non-fiction reading text. This focus on what real teachers and real students do in real classrooms in the Netherlands was chosen to ensure ecological validity. In terms of teaching style or pedagogical strategies the teachers were specifically asked to do what they always do. This is what is meant by the term “business as usual” lessons used in this dissertation, although we would like to indicate that teaching a lesson which is recorded on camera might never be completely business as usual.
It should be mentioned that data collection for this study finished four days before school closure due to the global outbreak of COVID-19. This means that this dissertation presents a detailed description of classroom practice in the Netherlands right before the unprecedented perturbation of the COVID-19 pandemic. At the time of writing it is unclear if and how the current state of the world might affect the ways in which teachers
and learners interact in physical or online language classrooms. Currently, in traditional media and on social media platforms, many teachers and learners express a longing for “real interaction” in a physical classroom. The general belief seems to be that when it comes to teaching, real interaction cannot be compensated for by online interaction. This claim, however, makes us wonder what some of the indispensable characteristics of real classroom interaction might be. The systematic descriptions presented in this dissertation may provide input for a reflection on “real interaction”.
6.4. Summary of the main findings
This dissertation provided descriptions of teacher-student interactions in business as usual reading lessons. The first study analyzed teaching strategies for scaffolding, the second study focused on micro-timescale of the lesson by zooming in on the dynamics of adaptive patterns formed by teacher questions and learner answers, the third study focused on patterns of stability in interaction by formed teacher questions and learner answers, and the fourth study used questionnaire data to validate the observational results.
6.4.1. Teaching strategies for scaffolding EFL reading comprehension (chapter 2)
In a multilingual world, critical literacy skills are important. The study presented in chapter 2 focused on strategies teachers use to foster learners’ reading comprehension. The study described how the advanced teaching skill of scaffolding can be operationalized in an event-sampling instrument suitable for observations of English as a foreign language (EFL) lessons in which teachers use a text as language input. The study builds on the scaffolding framework presented in the review study of Van de Pol et al. (2010). Based on an extensive analysis of the research literature, Van de Pol et al. (2010) found that there are six different ways for teachers to use scaffolding during classroom interaction:
asking questions, modelling and giving instructions, explanations, feedback and hints. The review defines scaffolding as providing calibrated and adaptive support and indicates that providing the right support depends on the teacher’s diagnosis of the learners needs, but also on ways in which teachers can make lesson content comprehensible for leaners (Van de Pol et al. 2010). The same review study also found an important limitation of the scaffolding research. Due to the complex and dynamic nature of the scaffolding construct, whether or not support provided by the teacher to the students during the learning process qualifies as adaptive, calibrated and contingent makes scaffolding behavior notoriously difficult to measure.
The measurement problem regarding scaffolding also exists in the field of second language teaching and learning, but is understudied. Existing observation protocols for the language classroom such as Motivational Orientation of Language Teaching (Dörnyei & Guillouteaux, 2008) and Sheltered Instruction Observation Protocol (Echevarria et al., 2010) do not address the complexity of the scaffolding construct. These studies assume that the presence or absence of teacher scaffolding in the language learning process can be reliably observed from one high inference observation point in which the term scaffolding is used. This, however, fails to do justice to the multifaceted nature of the construct, which was considered to be a gap in the existing research literature. In order to address the complexity aspect of measuring scaffolding, the English Reading
Comprehension Observation Protocol (ERCOP) was developed. In this observation protocol
teaching strategies for scaffolding English reading comprehension were operationalized. Based on the research literature for teaching foreign language reading comprehension and content-based language teaching a list of teaching strategies was composed. A core criterium for the selection of the strategies, was the teacher as an activator of cognitively challenging interaction. Teaching strategies were categorized into scales for modelling, feedback, instructions, explanations, hints and asking questions. These six categories were the scaffolding means as identified by Van de Pol et al. (2010). Next, a
long list of observation points was piloted and reduced to a list of items that caused little debate among the observers. This resulted in a list of observation points with teaching strategies for scaffolding English reading comprehension which was consequently used in a generalizability study.
A generalizability study can be used to disentangle possible sources of variance in the observed scores. Whenever the focus of an observation is on a complex construct such as scaffolding, the observed scores might be influenced by external factors such as different interpretations and observer beliefs about what the construct entails and how to link the construct to actual behavior. In any observational study it is important estimate the impact of who is looking, on which moment, and how often. In order to address possible interaction effects in the observed scores that might be caused by the teacher, the observer or the lesson, a generalizability study was carried out.
In this observational study the focus was on observing the range of teaching strategies used by a teacher in an EFL lesson. Generalizability theory (Shavelson & Webb, 1991) is a statistical framework building on the assumption that sources of variation between the observations might be caused by potential sources of error. In a generalizability study (G-study) the magnitude of the sources of variance can be disentangled, which allows us to estimate the influence of for instance the observer or lesson. An important limitation of this design lies in the assumption that an observation of the teacher’s behavior was a regarded as a random sample of the typical behavior of this particular teacher and that it can be assumed that the behavior (here: average range of teaching strategies) is in a steady state. This also implies that in this study it was assumed that any differences in the observed scores were not caused by systematic changes in the teacher, but might have be caused by inconsistencies in the observations. The follow-up studies presented in the other chapters of this dissertation, addressed this limitation and investigated stability and variability in the teachers’ behavior.
Based on the outcomes of the G-study four additional observation points causing inconsistencies in the observations were identified. These observation points were therefore dropped. The results of the G-study performed on 41-item ERCOP instrument revealed that differences in the use of teaching strategies could be largely attributed to differences in teacher behavior. A low proportion of the variance in the observations was caused interaction effects between teachers and observers. This stresses the need of combining the use of an observation tool like ERCOP with observation training. It is essential that observers who use ERCOP look through the lens of the instrument and do not let judgements based on personal beliefs about good EFL teaching influence their scores. Rater training is therefore an indispensable part of any observation system. Whenever high stakes decisions are based on lesson observations, it is important to know how many observers are needed for reliable observational results. Therefore, a decision study (D-study) was performed as a follow-up study. The D-study results suggest that ERCOP scaffolding strategies can be reliably observed with two well-trained observers.
Chapter 2 explored the complex nature of scaffolding by using traditional and static event-sampling observation methods. This chapter can be regarded as a first attempt to address the measurement problem in scaffolding research. The follow-up observational studies in chapters 3 and 4 addressed the dynamic nature of scaffolding and approached the same measurement problem from a different angle, namely from a complex dynamic systems perspective, which does not focus on the outcomes of teaching or learning, but on characteristics of the process itself.
6.4.2. Observing the dynamics of co-adaptation in EFL lessons (chapter 3)
Whereas chapter 2 addressed the multi-faceted nature of scaffolding by focusing on many different ways in which language teachers support their learners, chapter 3 addressed the dynamic and adaptive nature of scaffolding. The study adopted time-series observations of naturalistic English as a foreign language lessons and zoomed in on one specific type of teacher-student interaction. Asking questions is one of the ways in which teachers may scaffold language learning (Van de Pol et al., 2010) and in combination with student answers this is common verbal classroom behavior that can be observed relatively easily. The teacher-student interaction was therefore observed by zooming in on patterns emerging from teacher questions and learner answers in the process of scaffolding EFL reading comprehension. The aim of this study was to investigate how teachers and learners adapt their questions and responses to each other. This bidirectional process is often referred to as co-adaptation (Larsen-Freeman & Cameron, 2008). The study presents a second observation scheme that was developed in this research project, i.e. the Questions and Answers in English Language Teaching (QAELT) coding scheme. QAELT coding scheme can be used to systematically observe the dynamics of teacher questions and student answers on the micro-timescale of the lesson. QAELT identifies the quantity and level of teacher question and student answer interaction. The coding scheme uses two ordinal scales for teacher questions and student answers which measure levels of cognitive complexity. After coding, teacher questions and student answers are synchronized resulting in time-series data that were analyzed with the State Space Grid technique.
The time-series data of four teachers, who were all observed on four different occasions, were used in a multiple case study. The results revealed that the process of co-adaptation between the teachers’ questions and students’ answers frequently led to patterns that were characterized by closed teacher questions and short or even absent student
answers. The analyses also showed that teachers ask many questions and that for every observed teacher there was at least one lesson in which we found significant evidence for students who adapt the level of their answer to the level of the teacher questions. This suggests that teacher questions to drive the interaction and that student adapt to the teacher. However, for only one out of fifteen lessons a positive significant correlation was found between the level of a student answer and the follow-up teacher question. This suggests that most teachers do not adapt the follow-up move to what the students said previously. In only one of the observed lessons the process of co-adaptation between teacher and students led to optimal interaction patterns which were formed by a relatively large number of open-ended teacher questions combined with extended student turns.
The results of this study have contributed to a better understanding of the dynamic nature of adaptive teaching. Moment-to-moment observations of questions and answers can be used to help uncover fixed or variable patterns and identify opportunities for teaching routines that promote active student participation. This study also showed that co-adaptation in teacher-student interaction cannot be understood without looking at what happened before a certain learning opportunity and what happens after. Finally, the study demonstrates how micro-level observations of questions and answers might be used as a first diagnostics step for teachers, learners or researchers who would like to identify classroom routines or, alternatively, opportunities for improvement and learning. Observations of business as usual might be used as a springboard for classroom interventions aimed at increasing the quantity or quality of question and answer sequences.
6.4.3. Co-construction of meaning through teacher questions and student answers (chapter 4)
The aim of the study in chapter 4 was to investigate an outcome variable of successful scaffolding in the language classroom, namely co-construction of meaning. Co-construction of meaning (Council of Europe, 2018) might be viewed as a favorable pattern of co-adaptation between teacher and learners. Extending Dewaele’s orchestra metaphor (2020), a favorable pattern of co-adaptation in teacher-student interaction might sound like a symphony. However, the results of this study revealed a dominant pattern in which the metaphorically enthusiastic conductor seemed to invest a lot of energy in trying to get the musicians to do more than merely picking up their instruments in order to make an occasional staccato sound. In a music repetition (the lesson) this might be a very frustrating situation for both the conductor (the teacher) and the musicians (the learners). More specifically, looking at the empirical data that were collected in this study, we observed teachers who ask many questions which were left unanswered or were merely answered by student utterances consisting of one, two or three words.
In order to find out whether the observed patterns from the multiple case study in chapter 3 also occurred in a larger sample, 40 business as usual EFL lessons were observed with QAELT coding scheme. In this quantitative follow-up study, 2727 sequences of teacher questions and student answers were coded and analyzed. The analysis of the distribution of the number of teacher questions for every lesson in the sample roughly followed a normal distribution, with a mean number of teacher questions of more than one question per minute. These results support the idea that the teacher question is a very common and frequently occurring pedagogical tool. Similar to the results from chapter 3, the most frequently observed pattern in this study with a larger sample was also the closed question and simple student answer.
Next, we analyzed how patterns formed by teacher questions and student answer interactions might vary or stabilize during the lessons. Co-construction of meaning in this study was operationalized as extended episodes of open-ended teacher questions and longer student answers. In order to gauge how frequently interaction patterns formed by these combinations of questions and answers took place, and in order to find out how stable this type of interaction was, three zones of interaction with increasing levels of student activity were defined: zone 1 was characterized by minimal student activity and varying levels of teacher activity, zone 2 was defined by teacher questions students should reasonably be able to answer but to which students give minimal responses, and zone 3 was distinguished by active verbal behavior from both teacher and learners. The results revealed that most interaction in the observed lessons can be characterized by high levels of teacher activity and low levels of student activity, zone 1. Only one lesson interaction never had any interaction in this zone. In addition, a lot of teacher questions did not lead to increases in student activity and that teacher-student interaction was highly unstable. When a lesson enters any zone of interaction, interaction rarely remained in this zone more than two questions in a row. This was the case a large majority of the observed interactions. Only nine lessons had a slight but significant tendency to stay in a zone if interaction was already there. This suggests that teacher-student interaction which is formed by question and answer sequences is idiosyncratic and unstable. The results also revealed that the least stable type of interaction was interaction in zone 3, which is the type of interaction which is needed for co-construction of meaning. In this study, the emergence of co-construction of meaning would imply that an increase in the probability of staying in zone 3 if interaction was already there. This was not supported by the results. Teacher questions are common and frequent pedagogical behavior and are often aimed at eliciting answers and active engagement from students. The study showed that extended episodes of active engagement were scarce and highly idiosyncratic. This suggests that prolonged stays in zone 3, characterized by extended sequences of open-ended teacher questions without predefined student answers, might be teacher and
learner behavior to strive for. The volatile teacher-student interaction patterns observed in this study in which open questions play an important, but currently limited role, suggest room to find ways to optimize student contributions in order foster co-adaptation that leads to co-construction of meaning.
6.4.4. Teacher perceptions of observed interaction patterns (chapter 5) In an attempt to explore possible underlying intentions for the observed interaction patterns, a validation study was conducted. In this final study, a group of EFL teachers was asked to reflect on the observational results that had been collected in lessons taught by colleagues working in a similar context. Respondents were asked whether they recognized the observed patterns formed by a prevalence of closed teacher questions and simple student answers, and whether could point out possible underlying reasons and propose suggestions to improve classroom interaction. A large majority of the respondents in fact recognized the observed pattern of closed teacher questions and limited student responses and almost everyone thought improvement was possible. According to the respondents, emotional factors rather than students’ proficiency levels, lesson content, lesson activities or motivational aspects might explain limited student participation. Most of the respondents suggested to implement formative evaluation practices in order to make students feel more competent and some respondents suggested that interaction could be improved by using different teaching materials. As an additional area of investigation, respondents suggested to focus attention of ways to foster a safe learning environment. Classroom routines that are conducive to fostering engagement and active language use were mentioned as possible areas for future research. These results seem to support the idea that social, cognitive and affective elements play a role when a teacher wants to provide the right support at the right moment in time. In a language lesson there are always trade-offs between the available time (limited to 50 minutes), the number of students (often more than 25), the type of activities (whole class, group work or individual) and quality and quantity of feedback. Trying to meet all these
demands all of the time is an almost impossible challenge, which is why conscious but responsive and flexible behavior from both teacher and students is crucial in shaping the language learning process together.
6.5. Discussion and conclusion
6.5.1. Research methodological contributions
According to Mercer and Kostoulas (2018) research projects that put learners at the center could lead to a neglect of the role of the teacher. The opposite, however, a heavy focus on the teacher without paying enough attention to the learners might be equally problematic. Therefore, entering into a debate about whether research and practice should be primarily learner-centered or teacher-centered might not be very helpful in advancing our understanding of the language learning process. This dissertation has attempted to shift the focus to an instructed language learning-centered perspective (Larsen-Freeman & Cameron, 2008). In order to be able to investigate this process, methods for systematic classroom observations were developed. As we have seen in this dissertation, observation starts with determining who is observing what on which moment. In this dissertation we observed the complex and multi-faceted nature of scaffolding as well as the dynamics of scaffolding as an interactive process. This has resulted in the development of ERCOP and QAELT, two new instruments which can be used for classroom observational research. Both tools were developed for EFL lessons, but might also be used for observations in different teaching settings.
The application of a complex dynamic systems (CDS) framework to analyzing teacher-student interaction in the language classroom acknowledges the complex reality of language teaching and learning, but also enables us to be transparent about the decisions that were made in terms of data collection, variables, analyses, assumptions
and measures. In the present study, the CDS focus resulted in the decision to study teacher and learner behavior together. The CDS framework, however, has methodological implications, for instance in terms which measures allow for an objective, valid and suitable description of the process. Complex dynamic systems theory is a theory that focuses on the characteristics of change. In this dissertation complex dynamic systems concepts such as state space, co-construction, co-adaptation, self-organization, attractor states, stability and variability were operationalized and applied to teacher question and student answer data. Analyses of the teacher-student interaction process revealed that mechanisms such as co-adaptation or co-construction of meaning may emerge from self-organization, but also that the outcomes of this process cannot be simply attributed to either the teacher or the learner. In a language learning-centered approach, positive or negative outcomes of an interactive process cannot be simply attributed to either the teacher or the learners, but the interactive process is caused by the process itself. This phenomenon is often referred to as process causality or multi-causality. The construction of a coding scheme like QAELT which can be used in combination with the State Space Grid data visualization technique provides a methodological contribution to studying classroom interaction from a complex dynamic systems perspective.
6.5.2. Empirical contributions
The empirical contributions of the four classroom studies can be summarized as follows. First of all, teacher scaffolding in the language classroom can be best understood as a flexible behavioral repertoire. Secondly, observations revealed mainly patterns characterized by asymmetric adaptation, in which the learners tend to respond to the teacher but the teacher does not respond to the learners. Thirdly, teachers ask many questions, but these questions often lead to short student answers or even to no answers at all. Fourthly, teacher question and student answer patterns throughout the lesson are highly variable and co-construction of meaning emerging from open-ended teacher questions and extended student turns was not observed. Finally, teachers recognize the
dominant interactional patterns in which they are far more active than their students and attribute these patterns to emotional factors. Communicative language teaching methods are built on the assumption that using language to express yourself is crucial for language development. According to Mercer & Dörnyei (2020), language learning cannot take place without using the language, which foregrounds the fundamental need for deliberate and active student participation during the language lesson. Open-ended teacher questions aim to create a need for students to communicate, and would thus lead to active verbal student behavior. The results of this study revealed that the number of open-ended and complex teacher questions in the observed lessons was very low, suggesting that there is room for improvement in this area. Increasing the number of open teacher questions might be beneficial for fostering co-construction of meaning, but before trying to change anything, it might even be more important to observe the way students respond to teacher questions and vice versa. Observation of what is happening is an indispensable first step for anyone who wants to improve their understanding of the dynamics of co-construction.
6.5.3. Limitations and suggestions for future research
In a classroom-based research project from a complex dynamic systems perspective there are many possibly relevant aspects and variables that could have been studied. This dissertation reduced the art of language teaching to studying a very specific type of teacher-student interaction formed by question and answer sequences. There is no doubt that many other decisions about potentially relevant and reliable measures could have been made. Teacher questions, however, are common classroom behavior and provide plenty of opportunities for the students to deliberately practice speedy language retrieval. Moreover, a question can be viewed as pedagogical and verbal behavior that pushes for active language use (Mercer & Dörnyei, 2020). Typical characteristics of real-time interactions are real-time constraints, for instance limited processing real-time for students
after a teacher question. Future studies may want to explore the role of retrieval and processing time.
This dissertation focused on describing the role of teacher questions and student answers in fostering cognitively challenging interaction in the foreign language. An important limitation of this study is that the focus was on how a teacher interacts with a group of learners, whereas in many classrooms a lot of interaction happens between learners in small groups or pairs. Analyses of small-group interaction is likely to reveal different patterns of co-adaptation and co-construction. This was outside the scope of the present study. Future observational research might also want to zoom in on what happens when a student does not respond to a teacher question. It would be valuable to continue real-time interactional research and analyze the characteristics of teacher and learner behavior which can be observed when interaction stalls. Suggestions for observable behavior and possible measures that addresses this issue can be found in studies zooming in what learners learn from language lessons, which is often referred to as uptake (Nystrand et al., 2003). Existing uptake measures might be useful in evaluating the effects of scaffolding. A limitation of the present study was that the evaluation of the scaffolding effect was only analyzed indirectly. Existing coding schemes as described for instance in Alexander et al. (2017), might be adapted and combined with the dynamic systems from this dissertation. Analyses of teaching strategies following unanswered questions might focus on finding out whether strategies exist that are more effective than others in eliciting student answers in a follow-up move. Should effective teaching strategies be found in additional descriptive research studies, an experimental study might be designed in which teachers are trained in noticing and responding to inactive student behavior in order to increase students’ active verbal engagement.
The focus of the present study was on the micro-timescale of the lesson. An important assumption in complex dynamic systems theory is that timescales are nested.
Co-construction of meaning is a complex dynamic process, and in order to better understand how and when co-construction of meaning emerges, longitudinal studies incorporating different levels of granularity such as the meso-timescale of a school year or the macro-time scale of a few years are recommended. This would enable analyses in which interactions between the different timescales are possible. Future qualitative and quantitative observational studies into the dynamics of effective follow-up behaviors of teachers and learners might improve our understanding of how and when co-construction of meaning emerges.
Finally, the results from chapter 5 suggests that emotions and the affective side of teaching and learning play an important role according to teachers. Both teaching and language learning can be a stressful at times. Real-time verbal behavior in the classroom might be influenced by positive emotions such as enjoyment, curiosity, and interest, but also by negative emotions such as anxiety, embarrassment and boredom. Future studies into levels of active participation by learners and teachers during the lesson may want to aim for including dynamic measures that also address non-verbal teacher and learner behavior.
6.5.4 Recommendations for practice
Language is a beautiful, powerful but also complex and dynamic communicative tool. In this section, we address some of the implications for educational practice and policy that might be inferred from ecologically valid observational data gathered in a naturalistic classroom setting. This dissertation aimed to offer a research-based perspective to raise teacher and learner awareness of the shared responsibility in the process of language learning and teaching. The most important message of this dissertation for teachers, teacher educators, managers and policy makers is the idea that teachers and students shape the language learning process together and that knowledge about characteristics of the language learning process is beneficial for anyone who would like to improve
communication. In the current Dutch national curriculum, language (learning) awareness is an implicit goal. This dissertation has shown that analyzing behavioral patterns in verbal interaction between teachers and their learners can be an important first step in raising language learning awareness. Language learning awareness is a key concept in the recent proposals for the new Dutch curriculum (primary and secondary education) (Dönzelmann et al., 2020; Curriculum.nu, 2020). The proposals for modern foreign languages in the new Dutch curriculum express a need for explicitly incorporating language learning goals related to knowledge about language as a communicative system. This can for instance be done by incorporating attainment targets related to the extent to which students are fluent in verbalizing their thinking, or the extent to which verbal behavior of yourself and others leads to (mis)understanding other people’s intentions and perceptions.
According to Dewaele (2020) language teaching might rely more heavily on teacher-student interaction than some other subjects. An important factor in EFL classroom interaction is encouraging learners to use language meaningfully. Ideally, an EFL teacher would always be able to adapt the use of the foreign language in ways that challenge and encourage students to use this language as well. Additionally, this ideal EFL teacher would always know how to support students in order to overcome cognitive and affective roadblocks they run into when they are trying to verbalize what they would like to say. In the language learning process, language, for instance teacher questions, can be used as a pedagogical tool to foster for co-construction of meaning. This dissertation has shown that the ways in which teacher and students respond to each other from moment to moment are the building blocks for the process of interaction and more importantly, that there is a lot of room for improvement. Accurate and adequate descriptions of how teachers and learners respond to each other in business as usual lessons may form an invaluable starting point for professional development and might be used as a heuristic device, or as what Larsen-Freeman and Anderson (2011) refer to as a “foil for reflection”. Awareness often starts with observations of what is currently happening. In a simplified
form, the systematic observation tools developed in this dissertation might be used by teacher educators for professional development purposes in for instance video feedback coaching sessions (Menninga, 2017).
The teachers who participated in this study emphasized that connecting with students on both an emotional and a cognitively challenging level is very important. Systematic observation as a tool for language learning and teaching might be used to overcome strong convictions or unsubstantiated opinions by first directing the attention of the observers towards the more neutral focus on teacher and learner (verbal) behavior. This shifts the focus to what they are doing, not immediately to why people are doing whatever they are doing. When the learners’ responses are minimal, there are more options for teachers than merely moving on to the next question. Students could respond with a question themselves, but it is also possible for teachers to adapt their next turn to the previous student turn. This can be done by repeating a question or clarifying the teacher question which lead to a minimal student response. The process of co-construction of meaning is a joint process in which interactive turns can and need to be managed. Observations of a process might uncover implicit patterns, which might raise the awareness of the optimal and suboptimal patterns in teacher-student interaction.
A possible next step based on consensus of what is currently happening lies in a context-specific interpretation of what is going on. Contextual factors such as group size, curriculum and examinations might influence the extent to which classroom behavior is deemed effective or in effective. Accepting a shared responsibility for the learning process starts with explicit and objective answers to the questions: Where are we now? Where are we going in terms of learning outcomes? And what kind of action is needed from learners and teacher to get there? These three questions form the backbone of formative evaluation practices, which focuses on information that can be gained from a focus on the learning process rather than merely on the outcomes of learning and teaching
(Sluijsmans, 2020). Postponing the discussion about people’s underlying intentions by first focusing on the learning process may actually provide a neutral and safe stance for teachers and learners from which they are willing to accept the shared responsibility for getting it together.
