The project was financed by Erasmus+ KA220-SCH Cooperation partnerships in school education 2021-1-NL01-KA220-SCH-000032600: Engaging Languages in Intercultural Virtual Exchange (E-LIVE). We are grateful to the student teachers and our partner teacher-trainer who participated in the virtual exchange. This paper reflects the views only of the authors and the European Commission cannot be held responsible for any use that may be made of the information contained therein. Many thanks to the reviewers for their constructive feedback as well as to the editorial team.
1Telecollaboration or Virtual Exchange is becoming increasingly popular in pre-service teacher education programmes. O’Dowd (2018, p. 1) defines VE as “the engagement of groups of learners in online intercultural interactions and collaboration projects with partners from other cultural contexts or geographical locations as an integrated part of their educational programmes.” VE has several benefits often cited in the literature as learners’ opportunities to practise language and develop intercultural awareness and competence. Kleban and Bueno-Alastuey (2015) highlight that it also promotes critical aspects of pre-service teacher training including the development of language teaching skills relating to understanding and adapting texts and tasks, formulating instructions, and engaging in reflective practices” (in Emir & Yangin-Ekşi, 2024, p. 3).
2Wicking et al. (2021) report that the proliferation of technological tools and platforms enable teachers to design tasks which foster interaction and negotiation in the target language (L2). This places importance on understanding how to design tasks which promote language learning, intercultural awareness, and intercultural competence. Task-based language teaching (TBLT) centres on “the importance of engaging learners’ natural abilities for acquiring language incidentally through the performance of tasks that draw learners’ attention to form” (Ellis et al., 2019, p. i). TBLT’s emphasis on meaningful tasks can help teachers when designing VE tasks. While O’Dowd and Ware (2009) have described task types in VE including information exchange, comparison and analysis, and collaboration and product creation, others (Guilford, 1967; Udrya, 2021) have suggested that tasks which encourage multiple avenues of discussion (known as divergent tasks) and those which require learners to come to a consensus (convergent tasks) are also practical considerations in task design. Cappellini and Macré (2020) found that divergent tasks more easily align with the pedagogical objectives of a VE, while convergent tasks did not always support their task aims. They argued that greater autonomy in task design enables learners to expand on the topic, thus providing additional opportunities for learning about language and culture. O’Dowd (2016, p. 287) suggested reflection surrounding “the role of the medium in online communication” and “cultural self-reflection and critical evaluation” are essential in task design. Although several parameters for task design have been investigated, the need for pre-service teachers (hereafter student-teachers, STs) to be trained in TBLT task design for VE is critical. The present study seeks to fill this gap and provide recommendations to teachers and teacher-trainers regarding VE task design.
3Given the focus of the present study, investigating VE task design, we begin by reviewing a teaching approach often associated with virtual exchange: Task-based Language Teaching (TBLT). This approach exposes language learners to “real world” tasks which can be broken down into smaller tasks, i.e. a pre- and post-task, which take place before and after the main task respectively (Ellis, 2003, p. 16). In breaking down tasks into pre-, main, and post-tasks, STs are trained to consider all aspects of the lesson planning process.
4In line with calls for VE to be embedded into L2 curricula at all levels, González-Lloret and Ortega (2014) advocate for Technology-mediated task-based language teaching (TMTBLT), a curriculum which achieves this aim. Smith and González-Lloret (2021) argue that the role of technology is as important as the role language plays in tasks and outline key features of TMTBLT tasks as ones which are authentic to future activities of the learner and provide an element of reflection to encourage development.
5There are several reported benefits that teachers reap from teaching in a VE. O’Dowd and Dooly (2021) present these benefits as teachers extending their professional network, honing collaborative skills, and incorporating new teaching techniques and technology into their classroom. Teachers’ positive experiences when teaching VE include working and collaborating with their partner, feeling more capable of handling future teaching challenges, and matching new technologies to support learners’ diverse needs (Wu, 2023).
6Despite the reported benefits and calls for further integration of VE into teacher-training programmes, Hagley and Green (2022) report that few training opportunities exist for teachers in developing countries who cannot afford the cost of these programmes such as Unicollaboration. While resources are available1 on their website, these take the form of task descriptions that teachers would use without the guidance of a trainer which is fee-based. As they, and many others have suggested, for VE to become commonplace in the classroom, training is a necessity. This is evidenced by several studies which report that teaching VE requires a range of new skills and roles that teachers may not have learned in traditional teacher-training programmes such as organisational and technical skills (Turula & Raith, 2015), assistance with task design (Fuchs et al., 2017), and having constructive discussions with a VE partner (Helm, 2013). Further, Sauro (2016) finds that teachers who engage in VE, like their learners, likewise benefit from working with a VE partner when designing tasks.
7Tasks in language teaching and learning precede TBLT in general and VE in particular. Parameters that have been presented within the literature on VE task design (see Table 1) have been informed by such discussions around tasks. For example, Guilford (1967) outlines the difference between divergent and convergent tasks. An example of a divergent task may be discussion questions, while designing a holiday can be an example of a convergent task. Wicking et al. (2021) suggest that successful VEs follow the sequence of tasks set out by O’Dowd and Ware (2009). Within information exchange tasks, learners are typically sharing personal information about themselves as a means of an ice-breaking activity. This is followed by a comparison/analysis task wherein learners discuss and compare aspects of their cultures requiring in-depth thinking. The final task category pushes learners to think on the deepest level and negotiate a project outcome together. O’Dowd and Ware argue that adopting this sequence in VE task design builds complexity, both linguistically and reflectivity for learners. For additional information on strong and weak forms of VE task design, see O’Dowd (2016). Hoffstaedter and Kohn (2015) outline additional task design parameters for teachers to consider such as how learners will communicate with one another, the topics they will discuss, the language(s) used, their physical location during the exchange, format of the exchange, and technological tools being used. VEs often include both synchronous and asynchronous tasks as each have their strengths in the learning process.
8The aforementioned studies demonstrate critical decisions in task design that teachers should consider when designing VE tasks so as to foster learner engagement. Table 1 summaries these studies.
Table 1 – VE task design parameters.
Guilford (1967)
|
O’Dowd & Ware (2009, p. 176-77)
|
Hoffstaedter & Kohn (2015, p. 3)
|
-
Convergent (closed tasks)
-
Divergent (open tasks)
|
-
Information seeking
-
Comparison and analysis
-
Collaboration
|
-
Modes of communication (written & spoken); interpersonal; presentational
-
Topics: general, personal, specialised
-
Pedagogical language constellation: tandem; lingua franca
-
Physical locations: in-class from PC lab in school; outside-class from PC lab in school; outside-class from home
-
Blended learning ensemble: prep, main & follow-up; multimodal blend
-
VE tools: OpenSim, BBB; chat, forum
|
9Philp and Duchesne (2016, p. 3) define engagement as “a state of heightened attention and involvement, in which participation is reflected not only in the cognitive dimension, but in social, behavioural, and affective dimensions.” Studies have outlined teaching strategies which can promote learner engagement, such as illustrating how the learning outcomes are relevant to learners (Skinner & Pitzer, 2012) and providing learners with choices to engender learner agency (Reeve et al., 2004). Gijsen (2021) echoes this point, finding that part of the teacher’s role is to encourage learners to define task success personally and enable them to attain these goals autonomously.
10Despite the number of years that researchers have been conducting studies in learner engagement, there is little consensus on a definition of learner engagement (Christenson et al., 2012). Much of the research centres around the notion that learner engagement is comprised of mainly three dimensions: behavioural, what learners do in the task; attitudinal, how learners feel in the task; and cognitive, what learners think about the task (Fredrick et al., 2019; Christenson et al., 2012; Skinner & Pitzer, 2012). While the majority of the literature has investigated learner engagement in the traditional classroom (Finn, 1989; Pekrun & Linnenbrink-Garcia, 2012), an increasing number of researchers are interested in examining learner engagement in online learning contexts. For example, Gijsen’s (2021) doctoral work presents general and specific markers of task engagement to help teachers and researchers recognise how and whether learning is occurring in such environments. Gijsen’s seven general task markers of engagement include task management, topic and content, language, task appropriation, partner orientation, communicative participation, and technology. Table 2 presents the specific markers of learner engagement by type of learner effort (Gijsen, 2021, pp. 59-60).
Table 2 – Types of effort and specific markers of learner engagement.
Type of learner effort/engagement
|
Specific markers of learner engagement
|
Behavioural effort
|
-
Checking one’s technological infrastructure
-
Attending to the task instructions
-
Contributing to the discussion
-
Making arrangements for meeting each other
-
Prompting partners to say something
-
Reading the learning station boards
-
Responding to questions
|
Cognitive effort
|
-
Reflecting on understanding of the task
-
Reflecting on problems they face
-
Personalising the task
-
Asking questions
-
Giving explanations
-
Providing a simplified paraphrase
|
Attitudinal effort: Emotional
|
-
Liking the technology used
-
Liking the topic
-
Commenting introspectively (enjoying the task)
|
Attitudinal effort: Cognitive
|
-
Showing interest in the task
-
Commenting introspectively (feeling prepared for the task)
|
11In specifying general and specific markers of learner task engagement, Gijsen gives teachers a means of identifying these in their VE practices. She argues that while designing autonomy into online tasks can raise learners’ engagement, teachers must help learners to ascertain which parts of the task will need more of their focus. She, thus, recommends that teachers design “awareness-raising activities that help students to establish which requirements are relevant to pursue during an exchange are of crucial importance” (Gijsen, 2021, p. 146).
12Although the markers of learners’ behavioural effort is a topic worthy of discussion, the aim of our study focuses on the cognitive and attitudinal markers entailed in learner engagement which relate to our interest in learner agency and reflection in VE tasks. In doing so, we now present how learner and agentic engagement have been measured.
13The challenge of measuring the dimensions of learner engagement has been underlined in the literature. For example, researchers have sought to measure behavioural engagement by observing the amount of time a learner spends on the task (Philp & Duchesne, 2016). While self-reporting and interviews have been suggested as tools for measuring cognitive and attitudinal engagement (Philip & Duchesne, 2016), Dooly and Vinagre (2022) note that self-reporting lacks reliability as these instances are often too short, out of context, and censored so as not to hurt the teacher’s feelings (p. 399). To address this, the present study triangulates self-reported data with video interaction data of STs completing the VE tasks in real time. Ganassin et al. (2023) have proposed a ‘skills passport’ as an alternative means of assessment which tracks learners’ linguistic and cultural skills development in the VE.
14As previously mentioned, although most literature focuses on behavioural, attitudinal, and cognitive engagement, a fourth dimension of learner engagement has been introduced more recently: agentic engagement. This is defined as “what students say and do to create a more motivationally supportive learning environment for themselves” (Reeve & Shin, 2020, p. 151). In this manner, Reeve and Shin contend that agentic engagement is the sole dimension that is reciprocal in nature, as opposed to one in which learners passively accept knowledge. Simply put, in attempting to alter how they are taught, the learner impacts the teacher and vice versa. Pino-James et al. (2019) argue that agency is linked with autonomy: teachers who allow for more autonomy enable agency. Along these lines, teachers who design tasks that limit autonomy, likewise hinder agency resulting in disengagement in tasks (in Fredrick et al., 2019). As it may be difficult for teachers to determine when learners are being agentically engaged, Reeve and Shin (2020, p. 152) have outlined 12 functions and phrases that learners may use in class to demonstrate this proactive effort including telling the teacher what they are interested in learning, expressing a preference, and making a suggestion (see appendix A).
15Having discussed relevant studies in TBLT and VE teacher training, we highlighted several task design parameters which have informed our investigation into VE task design. With special interest in how STs show their engagement in online tasks, we propose two research questions for our study:
Research question 1: Between information-seeking, comparison and analysis, and collaboration VE tasks, is one task type reported as more engaging than others by student-teachers?
Research question 2: How do student-teachers demonstrate engagement in and reflection on the task reported as most engaging?
16This section outlines the participants, pedagogical design of tasks, as well as data collection and analysis methods. A VE between a Dutch and French university was conducted during the course of one academic semester (Fall 2022). Participants included 33 STs, 17 from a French university located in the central part of the country, and 16 STs from a Dutch University were also located in the central part of the Netherlands. French STs were STs completing their Master’s in Teaching, Education, and Training, specialising in English Language Teaching. Participants from the Dutch University, were completing their Master’s in Education. The aims of the VE included: (1) Knowledge of the modalities and specificities of teaching-learning through VE systems, and (2) Know how to implement a reflective approach to analyze your techno-semio pedagogical skills. The VE formed part of the Engaging Languages in Intercultural Virtual Exchange Erasmus+ project (E-LIVE, 2024).
17For the present study, STs completed three different task types which follow O’Dowd and Ware’s (2009) task types. Our first and second tasks were divergent with our final task being somewhat convergent. Hoffstaedter and Kohn’s (2015) parameters were agreed on by both VE partners at the start of the project. Table 3 outlines how these task design parameters were employed.
18It is worth noting that the VE required the STs to complete several parts autonomously including arranging times to meet their partner and recording the discussion on the videoconferencing platform BigBlueButton (BBB) as well as uploading a future VE task onto Moodle for their partner to see, and designing the final result of the task (following feedback from their partner) in Genially. All participants were provided the same task instructions on Moodle. The pedagogical design for the VE is outlined in Figure 1.
19The pre-task of task one was asynchronous and asked participants to record a brief introduction of themselves and post this onto the asynchronous video platform Flip (Microsoft, nd). The main task (1) required participants to comment on their partner’s introduction, while the post-task reflection asked participants to reflect on their level of satisfaction, communication, collaboration, intercultural experience, and the digital tools utilised. Please see Appendix C for the post-task instructions.
20The pre-task of task two included participants writing a forum post in Moodle describing their country’s education system and comparing their country’s qualifications for becoming an English teacher. The synchronous main task (2) was recorded by participants on BBB, with the post-task centering on responding to the same reflective questions as task one, with the expectation that participants would reply to relevant prompts.
21The pre-task of task three involved participants designing and posting on Moodle a VE activity they could potentially teach in the future. The focus of the final online meeting (task three) being that partners would provide feedback to one another’s activity, also on BBB. In this way, STs experienced the learner’s role of participating in a VE - including tasks they liked/disliked - as well as the teacher’s role of which tasks were designed more (or less) successfully. This knowledge could be applied to future VE task design.
22Due to differences in the partner institutions’ semester dates, the synchronous meeting for this task was reduced to a one-hour online meeting. The task was therefore designed so that this synchronous exchange focused on STs giving feedback on tasks designed individually during the pre-task phase rather than designing a task together.
Figure 1 – Pedagogical design of VE tasks.
23(Source of Figure 1: Wigham & Jauregi-Ondarra, 2023).
24Table 3 outlines how the tasks of the present study aligned with previous literature in VE task design provided earlier in Table 1. Gijsen’s (2021) behavioural, cognitive, and attitudinal efforts were incorporated into all three tasks and are therefore not added to the table.
Table 3 – Task type parameters of the current study.
Task parameters
|
Task one
|
Task two
|
Task three
|
Asynchronous
|
x
|
|
|
Synchronous
|
|
x
|
x
|
Convergent
|
|
|
x
|
Divergent
|
x
|
x
|
|
Information seeking
|
x
|
|
|
Comparison/Analysis
|
|
x
|
|
Collaboration
|
|
|
x
|
Mode of communication:
|
Presentational
|
Interpersonal
|
Interpersonal
|
Topic: General
|
x
|
x
|
x
|
Pedagogical language: ELF
|
x
|
x
|
x
|
Physical location: Outside class from home
|
x
|
x
|
x
|
Blended-learning: multimodal blend
|
x
|
x
|
x
|
Digital tools and platforms
|
Discussion forum (Moodle, Flip)
|
Video & chat (BBB)
|
Video & chat (Genially, BBB)
|
25Table 4 provides study participants’ demographic information. Participants’ names have been anonymised. STs were chosen based on varying degrees of participation, with pair 1 displaying the highest levels of class participation prior to the VE and pair three showing the lowest participation respectively. Further, it was decided early on to concentrate on the transcription of three pairs given time constraints. A parallel study regarding STs’ multimodality in the partner orientation task parameter was also conducted (Wigham, Jauregi-Ondarra & Bennett, 2024).
Table 4 – Participants’ background information.
Pair No.
|
Participant
|
Age
|
L1
|
1
|
Jennifer
|
21
|
French
|
1
|
Ettie
|
22
|
Dutch
|
2
|
Manon
|
-
|
French
|
2
|
Cora
|
-
|
Dutch
|
3
|
Cécile
|
-
|
French & Arabic
|
3
|
Maria
|
26
|
Dutch
|
26Data collection began in November 2022 following participants providing consent for their data to be used for research purposes. Ethics approval was granted to the Dutch University under application number FRTC-GE:23-090-02. As several data instruments were used to investigate the research questions including questionnaires, post-task reflections, and transcripts of synchronous task interactions, Table 5 outlines data extent and the word count by each data instrument. Although the post-VE questionnaire was the last data collected chronologically, it was our initial means of determining which task(s) STs reported as most engaging. This task was then verified against STs’ post-task reflections. Once these data instruments confirmed the task that STs reported as most engaging, the corresponding synchronous task was transcribed.
Table 5 – Data collected (n = 6 participants).
Post-VE questionnaire
(n = 5)
(written responses to open questions)
|
Post-task reflections (n = 13)
|
Task two Transcriptions
|
T1
|
T2
|
T3
|
Pair 1
|
Pair 2
|
Pair 3
|
Total word count
|
1,223
|
1,397
|
2,416
|
1,697
|
5,214
|
9,565
|
6,572
|
27The post-VE questionnaire contained 19 questions of which 11 were open and 8 were closed (see Appendix D). All but one of the closed questions provided a five-point Likert Scale for participants to report their response. Closed questions allowed for simple numerical analysis in the form of descriptive statistics while qualitative analysis of open-ended questions was completed thematically (Braun & Clarke, 2022) and agentically according to the table by Reeve and Shin (2020, see appendix A).
28To analyse the post-task reflective data, Bocquillon, Derobertmasure, and Demeuse’s (2017) levels of teachers’ reflective processes were employed. This reflective grid (see Table 6, inspired by Bocquillon et al., 2017, p. 11-12) helped to ascertain which tasks resulted in more depth of reflection by participants. Additionally, and to understand participants’ agentic engagement in the VE tasks, Reeve and Shin’s (2020) parameters of agentic engagement were applied to the dataset.
Table 6 – Reflective processes by level.
Reflective processes
|
Sample comment from teacher
|
Levels
|
Narrate/Describe
|
“Here, I ask them to analyze and synthesize the graph.”
|
1
|
Question
|
“I wonder if gapped texts have any interest.”
|
Become aware
|
“I realized that I had only spoken to a part of the class.”
|
Pointing out difficulties/problems
|
“I experience a lot of difficulties in grasping the essential elements of the content.”
|
Legitimising practice according to a preference, a tradition
|
"I have always enjoyed facilitating group work.”
|
Legitimising practice based on contextual arguments
|
“Such a method is not applicable given the number of students in the class.”
|
2
|
Legitimising practice according to pedagogical (P) or ethical (E) arguments
|
(P) “This method is more effective than another.”
(E) “Such an activity develops citizenship.”
|
Intentionalising practice
|
“My goal was to have them synthesize the material in their own words.”
|
Evaluating practice
|
“If I self-assess, I find that this activity did not work because the students did not understand the content hidden behind the activity.”
|
Diagnosing [practice]
|
“I didn’t consider this response because I hadn’t prepared my lesson sufficiently.”
|
Proposing one or more alternatives to practise
|
“Next time, I will ask another student to complete their peer's response instead of completing it myself every time.”
|
3
|
Exploring one or more alternatives to practise
|
“Next time, I will implement a technique to engage all students. Among the two techniques I have in mind (randomly questioning students and using individual whiteboards), I will choose the first one because it seems more suitable for the types of 'objectification' interventions I have planned for my next lesson.”
|
Theorising [different outcomes of practice]
|
“Thanks to this student comprehension issue, I discovered that one of the professional rules is to translate 'scholarly' material into accessible material for students.”
|
29Concerning the video recordings, the first stage consisted of transcription. Second, preliminarily coding was completed in alignment with Gijsen’s (2021) specific task parameters (table 2 above) for measuring task engagement in virtual tasks. Third, Grounded Theory analysis (Strauss & Corbin, 1998) was applied to identify potential new codes within our dataset which are an extension of Gijsen’s work.
30All analysis was preliminary completed by the first author, then discussed and agreed on with the second. Triangulation of sources was achieved in the comparison of data collected by these aforementioned data instruments to validate our findings.
31To answer the research questions, we present our findings by each research question. Responses from the post-VE questionnaire were analysed and are presented to answer the first research question: Between information-seeking, comparison and analysis and collaboration VE tasks, is one task type reported as more engaging than others by student-teachers?
32Numerical analysis of the closed questions in the post-VE questionnaire support that task two was the most engaging as shown in Table 7. Three STs reported feeling ‘engaged’ by task two and when asked to rate this task amongst the others, four found it to be ‘good’. In the larger dataset, 13 participants rated this task as ‘good’ with nine citing this task as ‘very good’. Five participants completed the questionnaire, raw numbers are provided in Table 7.
Table 7 – Participants’ self-reported engagement in VE tasks.
Questionnaire item:
|
Participant response:
|
How engaged did you feel in each task?
|
Task one
(Getting to know you)
|
n = 2 Engaged
n = 2 Somewhat engaged
|
Task two
(Education systems)
|
n = 3 Engaged
n = 2 Very engaged
|
Task three
(Task design)
|
n = 2 Very engaged
n = 2 Engaged
|
Please rate the tasks
|
Task one
|
n = 3 Neutral
|
Task two
|
n = 4 Good
|
Task three
|
n = 2 Good
n = 2 Neutral
|
33Concerning the thematic analysis performed on the open-ended questions of the post-VE questionnaire, three themes were identified. These themes included positive aspects that the STs felt they received from the VE, the challenges they experienced, and what they felt made the VE successful as shown in Table 8.
Table 8 – Themes from open questions in the post-VE questionnaire.
Theme
|
Participants’ comments
|
Counts
|
Theme 1: Positive aspects of the VE experience
|
Partner collaboration
|
8
|
Learning about another culture
|
5
|
Practising English as a Lingua Franca
|
3
|
Digital literacy development
|
2
|
Theme 2: Challenges experienced in the VE
|
Technical issues with tools (Flip, BBB)
|
9
|
Lack of clarity in carrying out tasks
|
4
|
French STs: Repetition in task design
|
2
|
Theme 3: Successful aspects of VE design
|
Personal development (reflection and task design)
|
2
|
Completing tasks that built in complexity
|
2
|
34Within the first theme, participants’ comments about the positive aspects of VE align with previously reported benefits in the literature. Notably, participants felt that the most positive aspect was collaborating with their partner, which also supports the numerical and reflective analysis that the synchronous tasks were more engaging than the asynchronous ones. Broadly, participants reported problems using the digital tools (Flip and BBB). More specifically, the French participants suggested that elements of the VE tasks were repetitive, while the Dutch participants felt the overall task design lacked some clarity. Regarding the aspects they felt that made the VE successful, comments centred on feeling that they were developing personally both in their teaching practice and designing future VE tasks and that tasks were well sequenced.
35Analysis of the transcripts of task recordings was employed to answer the second research question: How do STs demonstrate engagement in and reflection on the task reported as most engaging?
36To begin, we tallied the most frequent specific task parameters from behavioural, cognitive, and attitudinal engagement from Gijsen (2021) that were found in our dataset (Table 9).
Table 9 – Gijsen’s (2021) most frequently cited specific task parameters in task two.
Specific task parameter
|
Counts
|
Pair 1
|
Pair 2
|
Pair 3
|
Behavioural Engagement: Prompting partners to say something
|
12
|
10
|
8
|
Cognitive Engagement: Personalising the task
|
10
|
11
|
9
|
Attitudinal Engagement: Liking the topic
|
10
|
11
|
13
|
37In terms of specific markers depicting behavioural effort by participants, the most frequently cited marker was ‘prompting partners to say something’. Being an active listener was apparent in that STs often asked questions based on what their partner had previously said. The most frequent specific marker within cognitive engagement was what Gijsen (2021) referred to as ‘personalising the task’ which we interpreted as any attempt by participants to change or ask additional questions based on information their partner had provided. For example, Ettie asked Jennifer if she was drawn to teaching because she enjoyed the subject (English) or teaching students and Céline asked Maria if she had a job before studying to teach English. These specific questions demonstrated to their partner they were listening and interested to learn about their thoughts and experiences. It also evidences their attitudinal engagement, by expressing their enjoyment in the task topic. Other expressions that were noted in the transcripts and which indicated heightened attitudinal engagement were laughter and smiling.
38STs’ post-task reflections were analysed according to the reflective processes of novice teachers outlined in Bocquillon, Derobertmasure, and Demeuse (2017) and presented in Table 6 above. Table 10 shows that participants demonstrated the most reflective processes following the second task with 48 references.
Table 10 – Reflective processes identified per task.
Task No.
|
Reflective processes & count
|
Totals
|
Level 1
|
Level 2
|
Level 3
|
Task one
|
20
|
13
|
6
|
39
|
Task two
|
18
|
20
|
10
|
48
|
Task three
|
17
|
17
|
7
|
41
|
Totals
|
55
|
50
|
23
|
|
39However, when the level of reflection is taken into consideration, most of this reflection is at the lowest level with 55 references. Table 11 provides an overview of the reflective processes identified by level and task. Only processes which were identified are listed. Within the first and third tasks, the type of reflective process described most in participants’ post-task written reflections included narrating or describing what they did in the task. Within the second task, ‘evaluating practice’ was the type of reflective process noted most by participants.
40Analysis of the post-task reflections indicates that while task two garnered the most reflection, when taking a closer look at the type of reflection completed, participants did not report significant cognitive engagement (level 3 processes are outlined in Table 6 above).
Table 11 – Reflective process by task.
Level of reflection
|
Type of Reflective process
|
Sample descriptors from our dataset
|
Task one
|
Task two
|
Task three
|
1
|
Narrate/Describe
|
I am liked to watch my teammate video as it is nice first encounter.
|
12
|
10
|
8
|
Identifying difficulties/problems
|
My partner was late with uploading her task idea.
|
6
|
8
|
1
|
Legitimising practice according to a preference, a tradition
|
I think that intercultural mismatches were overcome thanks to the call that enabled us to clarify certain points.
|
2
|
1
|
0
|
2
|
Legitimising practice based on contextual arguments
|
It was easy to use and it was great that a meeting was already set up in the Moodle so I just had to join.
|
1
|
2
|
2
|
Intentionalising practice
|
I could make myself understood in the short time available as one of the French students responded to my video with some questions.
|
1
|
2
|
0
|
Legitimising practice according to pedagogical or ethical arguments
|
Her feedback was really detailed and precise. It was very useful to me.
|
2
|
0
|
4
|
Evaluating practice
|
I thought this tool was very suitable and convenient in use. It was easy to see the videos of the others.
|
8
|
14
|
5
|
Diagnosing
|
My partner was more focused on the knowledge a teacher should have about the English language instead of … classroom management or pedagogical skills.
|
0
|
1
|
0
|
3
|
Theorising [different outcomes of practice]
|
I could look up information about the French school system beforehand and prepare specific questions.
|
1
|
1
|
1
|
Proposing one or more alternative solutions
|
Maybe it would have been better if we had created the task together.
|
5
|
9
|
4
|
41To understand whether participants would want to make changes to the tasks, their post-task reflections were also analysed according to the agentic functions outlined in Reeve and Shin (2020, p. 152). Only the agentic functions that were found in the dataset are provided in Table 12.
Table 12 – Agentic functions in Reeve and Shin (2020, p. 152).
Agentic function
|
Task one
|
Task two
|
Task three
|
Let the teacher know what you are interested in.
|
0
|
1
|
0
|
Make a suggestion
|
1
|
0
|
0
|
Offer a recommendation
|
3
|
2
|
6
|
42Table 12 shows that the most frequently used agentic function by participants was offering a recommendation which mainly centred on requesting more time to complete the task. For example, Ettie wrote:
It was a pity we only had three minutes [to introduce ourselves], so next time I could ask whether it would be possible to [have] extra time on Flip....
43Another request by participants was that more clarity was needed in completing the tasks. For instance, Maria wrote about the confusion she experienced in the collaborative task (task three):
I thought the task was a bit confusing. I had to give written feedback to my partner and then she had to just take a look at my idea and give me feedback during our video call. At least, this is what I understood. I’m not completely sure if we executed this correctly.
44Here the slight difference in what the Dutch and French participants had to prepare before meeting online caused Maria to be uncertain of what she needed to do in the task. Scheduling issues between VE partners is an often-cited complication when designing VE, and was the reason for a slight difference in task instructions, though much research argues the benefits outweigh these issues (Emir & Yangin-Ekşi, 2024).
45Jennifer also commented on a change to task two's design which she thought would make it more interesting centering on matching STs who have similar levels of teaching experience:
Maybe for the next time, what would be interesting is to have a partner at the same level as us, that is to say a future teacher.
46While her partner Ettie provided the only suggestion for altering task one’s digital tool and instructions:
If I could change the digital tool (Flip in this case), I would add subtitles underneath all videos in which you could enter your name, home country and university. I think that would be useful.
47While there was not a great variety of agentic functions in the post-reflective tasks, this is perhaps not surprising given the mode of reflection being completed individually (self-reporting).
48While analysis of the reflective processes aligned with analysis of the post-reflective tasks, showing that participants mainly reflected at level 1 (14 instances), many participants identified difficulties or problems (12). Level 2 was the second most cited type of reflection (12 instances) mainly comprising two types: legitimising practice based on pedagogical arguments (5) and intentionalising practice (4).
49In response to RQ1, participants reported task two as being more engaging of the three tasks. This was detailed in participants’ responses to the post-VE questionnaire wherein 4 (out of 6) rated the comparison and analysis task as ‘good’ and more reflection was demonstrated in task two. Our agentic analysis showed that participants would change task three to be more like the second task reinforcing their preference for this task type. Hence, we examined the synchronous video recordings of the task two interactions to identify how participants’ demonstrated reflection and engagement in this task (RQ2). We believe that participants reported task two as engaging because they were agentically engaged with the task by altering it to share more personal information about one another. This is ascertained by comparing the online interactions with the instructions that were provided to the STs at the start of the VE which did not require STs to share personal details (appendix B). Grounded theory analysis evidenced participants using life experiences as examples during their interactions, with pair two personalising the topic the most. It also helped identify recommendations for future teacher-trainers in VE task design as outlined in the following section.
50We now discuss our findings in relation to the research questions.
51Research question 1: Between information-seeking, comparison and analysis, and collaboration VE tasks, is one task type reported as more engaging than others by student-teachers?
52Data from the post-task reflections and post-VE questionnaire show that STs reported task two (synchronous, divergent, comparison and analysis) as being the most engaging. STs cited they struggled most with learning to use the technology in the VE, namely Flip (task one) and BBB (tasks 2 and 3), with many of their comments relating to feeling stressed with the uncertainty of whether their videos had been recorded in BBB following initial issues with uploading their introductory videos onto Flip in task one. Several studies have outlined the need for learners to learn how to use new technological tools and/or platforms in advance of using them (Wicking et al., 2021). Within the present study, STs were given time to explore the tools individually in class, under teacher guidance, as a means of fostering autonomy.
53Despite challenges with using the digital tools in the tasks, it appears that this did not greatly impact STs’ positive perception of the VE tasks. Many STs reported they enjoyed collaborating with their partner and learning about their partner’s culture as the most significant advantages. These findings align with much of the literature on teachers’ reported benefits of participating in VEs (e.g. Wu, 2023). The comparison and analysis task also created more avenues for reflection, with many STs evaluating what they learned in the exchange about the partner’s education system and qualifications of becoming an English teacher against their teaching practice. Often, these comparisons were complementary, such as Manon who acknowledged that having different curricula for mixed-level classes could help learners who were performing well or struggling. STs also described what they had discussed with their partner in their post-task reflections, suggesting that the instructions of the post-task reflections could be modified to encourage deeper levels of reflection. Researchers have cited reflection as a critical part in VE task design (Smith & González-Lloret, 2021).
54STs’ post-task reflections were also analysed for agentic engagement, with STs suggesting the most modifications to the collaborative task (task three) in our study. This is interesting as their suggested changes were to redesign task three so that it was more similar to task two, which further demonstrated their preference towards task two as being the most engaging. Having STs participate in VE tasks as learners themselves provides them the unique opportunity to experience VE both as a learner and as a teacher. This also aligns with the criteria we outlined earlier for TBLT tasks being meaningful and connected to real world tasks that STs would have. In other words, STs may have reported task three, planning a future VE task for their learners, as more engaging had it been more collaborative in nature. The difference in academic semesters between the French and Dutch universities meant that STs could not co-create a VE task. This result was taken into consideration in the 2023-24 instantiation of the VE which opens perspectives for a future replication study.
55Research question 2: How do student-teachers demonstrate engagement in and reflection on the task reported as most engaging?
56Following STs’ self-reporting that task two was most engaging, we delved deeper into the transcripts of this interaction applying grounded theory analysis. In comparing our data to Gijsen’s (2021) specific task parameters, we extend her cognitive engagement parameter related to task personalisation to include more positive connotations which we call ‘sharing personal details or previous life experiences’. Regarding the first pair, Jennifer and Ettie shared personal stories about their mothers being teachers as inspirations of becoming an English teacher, the difficulties of being a student during the pandemic and what motivates them to teach. Along this vein, Manon and Cora asked one another personal questions about which English accents were difficult to understand in the United Kingdom and the (unfavourable) teaching conditions in their countries. The third pair, Céline and Maria, talked about the importance of maintaining mental health as a teacher. The range and depth of these topics demonstrate their engagement in the comparison/analysis task.
57Concerning attitudinal engagement, we noted that STs appeared quite comfortable in the online environment, further supporting the rationale for asking personal questions and sharing life experiences. It is unlikely that our STs would have discussed topics such as mental health/burnout as teachers and challenging accents as these are not topics that the trainer can anticipate. With STs taking control of the task and making it their own, we can see their depth of engagement. This engagement could also be explained as many STs were of approximately the same age and had grown up using technology. Another possibility could be that in utilising some of the task parameters outlined by Hoffstaedter and Kohn (2015), for example learners speaking with their VE partner from their home, the topic of talking about general differences between their education systems could have also impacted their level of comfort in the interactions we analysed. Our findings echo previous research by Cappellini and Macré (2020) that learners who modify task instructions to ask new questions are more engaged emotionally. STs’ emotions are evident in their write-in comments of the post-VE questionnaire wherein the most adjectives were positive including ‘comfortable’, ‘fine’, ‘easy’, ‘interesting’. Dooly (2011, p. 72) also finds that learners who modify task instructions are demonstrating being “active agents in the learning process” and by doing so are more engaged in the task.
58The analysis of STs’ post-task reflections shows that in speaking with their VE partner, who was also training to become an English teacher in another country, helped them to evaluate the VE tasks. In evaluating the tasks, they were also reflecting on future VE tasks in their classroom, as suggested within the literature. For example, all but one ST mentioned they had met the goals of the task, suggesting that STs were keeping in mind the learning objectives of the task and whether they felt they had met them. These results align with earlier research (Leone et al., 2023). STs also reflected on sharing the talking time during their interaction equally, showing an awareness of teacher talk time. Maria and Ettie commented that they felt the meetings with their French partners had gone well because each felt they shared both speaking and listening time. Although other STs’ reflective evaluations centred on their opinions of the level of communication, collaboration, and the digital tools used in the task, Maria in particular explained why she enjoyed the comparison and analysis task:
I found it enjoyable and informative to speak to another English teacher from another country. I was very interested in what she was talking about and I think she felt the same way towards my story.
59However, our analysis also shows minimal reflection by STs in the designed VE tasks leaving room for improvement. Indeed, our STs reflected more in the divergent tasks (one and two) as opposed to the final collaborative (convergent) task. Researchers recommend problem-solving/collaborative tasks as essential in VE as they create the conditions for more language use and negotiation (O’Dowd & Ware, 2009; Wicking et al., 2021). The fact that STs in our VE did not collaborate in the final task due to the schedule conflicts between the two universities is shown here in their lack of reflection.
60Reflection is a skill identified by the Council of Europe (2018) as one of the eight key skills that member states should develop amongst individuals. In teaching, this ability to stand back from one’s own actions (Forest & Lamarre, 2009) relates to a meta-competence which also facilitates the development of other skills (Ivanova-Armeykoya, 2020; Metaeva, 2006 cited in Ivanonva-Armeykoya, 2020) when articulated with authentic professional practices (Schneuwly, 2015). For this reason, considering how to develop deeper levels of reflection in future VEs is critical. One potential reason for the lack of reflection could also be the numerous reflective prompts we designed (see appendix C). Emir and Yangin-Ekşi (2024, p. 18) advocate that pedagogical mentoring is integral in successful VE task design especially in helping learners to create “appropriate discussion questions, encourage peer feedback, and allocate sufficient time” to complete tasks.
61Based on our study, we provide several recommendations for future teachers and teacher-trainers in VE task design. For example, we underscore that task personalisation is a critical element of learner engagement (O’Dowd, 2021), and one that can be easily built into task design. O’Dowd and Ware (2009, p. 184) have called this “tuning in” and suggest that learners connecting with their partner on a common interest is an indicator of a successful VE. Therefore, teacher-trainers must think and plan question prompts carefully in VE task design. As earlier research has suggested (Fredricks et al., 2019; Kurek & Müller-Hartmann, 2019; Gutiérrez et al., 2023), the teachers’ role in creating instructions which allow for autonomy and agency are essential. As recommended by Reeve et al. (2004), task instructions which provide choices can enable learners to feel more ownership when completing the task which allows them to define their version of task success personally. Emir and Yangin-Ekşi (2024) in particular, recommend formulating questions for discussion activities to include follow-up questions which focus on a cultural aspect their partner has just mentioned and follow-up questions which ask their partner to talk about a personal experience.
62Along the lines of task instructions, we recommend tailoring the instructions to promote deeper levels of reflection. Our post-task analysis shows that STs did not reach the deepest level of reflective processes outlined by Bocquillon, Derobertmasure, and Demeuse (2017). This can be remedied by designing questions which foster in-depth reflection as recommended by Brooke (2012). For example, the second level of reflective processes, “Evaluating practice”, asks teachers to consider whether planned activities had achieved their intended goals. An adapted reflective prompt could be: Were the aims of the lesson achieved? If so, what did the students do to demonstrate this? Reflective prompts for level three’s, “Exploring one or more alternatives to practise”, could be designed to ask: Consider the type of activity you taught in class, is there another teaching technique that you could have used but didn’t? Would the results have been different? If so, how?
63One of the French STs mentioned the amount of writing in the VE to be repetitive and time-consuming. Tailoring and minimising the number of question prompts in the post-task reflections can also be a recommendation moving forward. In the present study, our post-reflective question prompts totalled 24 which aimed at promoting reflection retrospectively and two questions which asked STs to think about what they would do differently in the future (see appendix C). In listing multiple questions, we perhaps provided too much autonomy to STs, which resulted in STs answering the same (descriptive) questions rather than engaging more cognitively with the task. We suggest limiting the number of reflective questions to two or three per reflective process and also adjusting these according to the task type so that STs are making connections of the task with their teaching practice. This was trialled in the 2024-25 instantiation of the VE.
64Agentically-speaking, data from the post-task reflections and open-ended questions of the post-VE questionnaire provide several recommendations for VE task design. Most notably, STs recommended changes to the third task so that it would become more collaborative in nature which was incorporated in the 2023-24 instantiation of the VE. STs also requested more time be allotted for task one, mentioning the Flip tool itself could perhaps be replaced. In the 2024-25 instantiation we therefore trialled Padlet (Wallwisher Inc., n.d.). In sharing these modifications, STs are taking ownership of their learning and proactively creating an environment they find more engaging for learning. Wicking et al. (2021, p. 16) note, “It is important for instructors and students to be familiar with the technology being used.” Researchers have suggested tools such as Microsoft Teams (Microsoft, nd) and WhatsApp (Koum & Action, n.d.) could be easier for learners to use pedagogically. However, studies have also highlighted the benefit of students learning to use new technological tools (Zhang & Zhou, 2023). To be certain, there must be a careful balance when choosing the digital tool for teaching purposes. A recommendation would be to ensure that teacher-trainers have demonstrated how the tool can be used prior to STs using it in the task to minimise stress and confusion which negatively impact the learning process and task engagement (Philp & Duchesne, 2016).
65Our thematic analysis has highlighted similar aspects of attitudinal engagement as reported in the literature. For example, STs reported feeling comfortable and nervous when completing the tasks, providing practical suggestions for improving future VE tasks with STs. The suggestions will likely be ones they bear in mind when designing their own VE tasks for their classroom. Designing a VE is not an easy feat, as Sadler (2018, p. 220) cautions: “Be ready for the workload.” Yet there is a wealth of research available to help teachers begin this process. For example, Wicking et al. (2021) have assembled criteria of successful VEs which can be a welcome resource. We, therefore, believe our findings add to the existing research in designing VE tasks that promote learner engagement, agency, and reflection.
66It should be noted the scope of the research is relatively small, with our findings being based on data collected from six participants as part of a larger group of approximately 30 STs. While our discussion has aimed to generalise and compare these findings to the literature, it is limited in scope. Similar to Wu (2023), our focus on STs’ engagement in VE tasks tells a limited story of whether teachers can transfer these lessons into their teaching practice. While our study outlines how future educators can apply our methodology to a larger cohort of trainee teachers, there remains the challenge of obtaining consent by participants in the VE.
67It is worth noting that the reflective grid we employed to analyse the post-task reflections is better suited towards STs who have begun teaching placements. While the Dutch STs had teaching placement experience to reflect upon, their French counterparts had only observed lessons in schools. The French STs showed higher levels of agentic engagement in their post-task reflections than their Dutch counterparts which is why both of these analyses (reflective and agentic) were conducted.
68Our study has investigated learner engagement across three different task types: exchanging information, comparison and analysis, and collaboration. We also investigated other task parameters by Guilford (1967) and Hoffstaedter and Kohn (2015) to understand which task design elements were reported by our STs as more engaging. To determine which VE design parameters were optimal in our study, we analysed STs’ comments regarding which task they felt more engaged by and cross-referenced this against their post-task reflections. Having established that STs found the second task (synchronous, divergent, comparison and analysis) as the most engaging, we took a closer look at the transcripts of this interaction. In employing grounded theory analysis and comparing our dataset against other work in online learner engagement (Gijsen, 2021), we report STs felt engaged in this task because they were able to ask new questions which enabled them to share personal details or experiences. Indeed, the questions being created by learners is a takeaway of our study and one that has been suggested in earlier research (Lambert & Zhang, 2019).
69Analysis of STs’ agentic engagement from post-task reflections demonstrates that changes STs would make to the task design included making alterations to task three so it would be more similar to task two, underscoring their engagement in the comparison and analysis task. While STs also reflected more in task two than the previous and following task, a closer look at these reflective processes highlights that the STs did not cognitively engage with the task in depth. Therefore, we have outlined several recommendations in task design instructions: first, instructions for synchronous tasks should invite STs to share personal information as it relates to the topic of discussion via examples of experiences. Second, limiting the number of reflective questions and ensuring these questions span reflective processes will enable STs to be more engaged cognitively.
70The findings of this study underscore the complex decisions that are made in designing VE tasks. The lessons that were learned from this VE included adapting the final task to be more collaborative, a change that was made following data analysis for the subsequent VE iteration. STs also commented that VE tasks were ‘time-consuming’ at times, suggesting that some of the questions in the pre/post tasks could be minimised. In providing feedback as learners in the VE, they learn which VE tasks, platforms, and tools are successful when designing their own VE tasks for future learners. For VE to continue to grow in mainstream teaching practice, studies that report on VE task design are critical. Indeed, we hope our study has contributed to a growing body of literature that can stimulate teacher and teacher-trainer reflection on the design of VE tasks.