2.1 Participants

The present study was carried out at The Open University (OU), a large open and distance learning provider, where language courses are designed with the latest theoretical approaches and research-based pedagogies for online language teaching in mind, thus potential challenges are addressed proactively. The language modules are taught by combining asynchronous activities with optional synchronous activities. The asynchronous component involves students working through online activities according to a study planner, and these activities include both independent work and opportunities to engage with peers through the use of forums, wikis, glossaries, and personal blogs. Each student is allocated a personal tutor who supports their study of the materials and marks and gives feedback on assignments. Students are encouraged to meet in online rooms as well as to attend tutor-led online group tutorials. Establishing enduring connections is challenging, as students can attend any tutorials they choose, so they may not see the same fellow students or tutor each time. Furthermore, participation in tutorials is advisory, not mandatory, so students might choose not to attend any synchronous sessions at all.

Participants were 307 OU undergraduate students in the School of Languages and Applied Linguistics (LAL), aged 18 and over. Most students study part-time and are in employment. To comply with research ethics and data protection, students were provided with an outline of the study prior to data collection with the approval of the OU Research Ethics Committee (HREC/4164). They were assured that their identity and all data obtained would be kept strictly confidential and solely used for the purpose of this study.

2.2 Instruments of data collection

For the study, a mixed methodology was used in sequential stages. First, a questionnaire with closed and open questions was sent to 999 OU language students in December 2021 to understand students’ views on FLSA, identify the main causes of it in online speaking situations, and to discern students’ speaking practices. Qualitative data were obtained from open-ended questions on the questionnaire and quantitative data from closed questions, with a mix of single and multiple-choice answers. In the questionnaire design, we drew upon data from two earlier events in LAL, attended by 88 students, to discuss how they encounter FLSA. As the questionnaire took approximately 20 to 30 minutes to complete, three £10 Amazon vouchers per language (Spanish, French, German) were raffled among students to incentivise participation. Of the 999 students who were invited to complete the questionnaire, 307 replied. The notable response rate of 30% underscores the significance of the topic for language learners. As part of the questionnaire, students could opt into further research via four group interviews (GIs) of approximately 60 minutes each with a view to gaining deeper insights into students’ fears and coping strategies.

This sequential mixed-method design was used for both explanatory and exploratory purposes: The qualitative data obtained from the GI sessions provided more explanation for some of the questionnaire data and analysis and allowed participants to explore and test mitigating strategies. Participation in GI sessions was incentivised via the offer of £10 Amazon vouchers. The GIs were conducted between February and June 2022 in English only and were conceptualised as interventions in the sense that they were thematic where coping strategies were elicited and discussed, and students could then apply these to their upcoming tutorials and discuss their impact in subsequent GI sessions. The topics discussed in the four GIs were (1) understanding foreign language speaking anxiety and how it affects the participants, (2) positive ways forward and what strategies participants have tried, (3) learning environments, and (4) future study plans. Tutors facilitated but did not influence the discussions, encouraging students to share experiences and strategies and give advice to each other. Of the 108 students who applied to take part in GI sessions, 10 were selected. Selection criteria included an even spread across the three languages (Spanish, French, German), levels of languages (beginners’ level/level 1, level 2, and level 3), and even distribution of students who declared additional requirements, those who declared FLSA, and those who did not. The four GI sessions were moderated by three tutors who were briefed so that they could facilitate discussions and approaches to mitigating FLSA.

2.3 Data analysis

Quantitative data focused on participants’ synchronous and asynchronous online speaking practices, their feelings towards these, and potential FLSA mitigating strategies. All quantitative data were transferred to Excel to carry out calculations.

Qualitative data focused on exploration and testing of mitigating strategies of FLSA. The four GI sessions were transcribed to undergo template analysis, together with the qualitative questionnaire data. Template analysis is a form of “thematic analysis which emphasises the use of hierarchical coding but balances a relatively high degree of structure in the process of analysing textual data with the flexibility to adapt it to the needs of a particular study” (Brooks, McCluskey, Turley & King, Reference Brooks, McCluskey, Turley and King2015: 203). This method of data analysis was chosen for its ability to accommodate multiple interpretations (Madill, Jordan & Shirley, Reference Madill, Jordan and Shirley2000) and predefine themes relevant to the study, which ensures consistency and rigour in data interpretation. We followed the six procedural steps recommended by King (Reference King, Symon and Cassell2012): familiarisation with the data sets to be analysed, preliminary coding, organisation of emerging themes, development of an initial coding template, its application, and finalisation. All coding was carried out in NVivo 12.

3.1 How do students encounter FLSA in the online classroom?

The first research question (RQ1) asked how students encounter FLSA in the synchronous online classroom. As shown in Figure 1, many of the perceived worries in the online classroom are very similar to those attested in the literature on FLSA in f2f classrooms, like general apprehension and lack of confidence when having to speak in the target language (TL), fear of having to speak spontaneously, fear of being judged, or perceiving that other students are more competent in the TL. While only a quarter of our students feel confident when having to speak, the same proportion of students are too anxious to interact and participate in the sessions, and one fifth choose not to attend at all because of their FLSA. Almost half of our participants are self-conscious about speaking in front of their peers and the tutor, and 38% panic when put on the spot.

Figure 1.

“Choose as many statements with regard to your speaking in online tutorials as apply” (multiple answers allowed; 305 students answered).

3.1.2 Breakout rooms

Tutors frequently employ breakout rooms to encourage collaborative work online in small groups, provide peer-to-peer contact, and a safe space to maximise speaking opportunities for each student. Our data suggest that in the synchronous online context, one of the main triggers of FLSA are precisely breakout rooms. Close to 60% of our participants feel anxious about group activities in online tutorials (Figure 2).

Figure 2.

“Rank the following statement about your speaking practice: I feel anxious about group activities” (single choice, 307 answers)`.

Some say they do not attend tutorials because they don’t want to work “in small groups with strangers”. The feeling of others being more competent in the TL is also magnified in this context, as attested by these accounts:

“I feel as if […] at least half the group have lived [in] the target language country. It is […] very demoralizing for those who haven’t.”

“If I’m in a breakout room with somebody who can speak exceptionally well […] and it’s their mother language […] I just get left behind and lost and I don’t particularly want to get involved as much either, I feel quite isolated quite quickly …”

When collaborating in small groups, some students fear not being able to understand what others say – that is, that their language level is inferior to that of their peers: “… there was an awful lot of new vocabulary so I couldn’t understand very much about what was being said and that made me feel pretty rubbish”. Another recurring issue is turn-taking in breakout rooms. On the one hand, students might perceive that someone dominates the conversation, or, on the other hand, that no one wants to speak or that the discussion simply does not flow:

“It’s stressful when you go into a breakout room and someone takes over the conversation and doesn’t allow time to reply.”

“I feel that if I volunteer to speak too often, I will seem to be ‘hogging’ airtime.”

“… there’s always like a little virtual dance almost of who’s going to be the first one to speak and bite the bullet and practise first.”

Although breakout rooms are not always perceived as negative, they seem to magnify emotions: “Breakout rooms can be either fantastic or awful, depending on who’s in the breakout room with me.” Some might feel relieved in the breakout room as they think they can ask for extra clarification, which they were too shy to do in the main room:

“[…] we go into a breakout room and I relax a bit because I confess to my collaborators that actually, I didn’t understand what are we doing and can I speak in English please.”

“I found that I, it was really good for me because I realised I wasn’t alone in this awful place.”

3.2 What strategies do fully online language learners apply to overcome FLSA in synchronous tutorials?

Figure 3 gives a general overview of the strategies employed by the participants of this study to manage FLSA (RQ2). Although most of our participants experience FLSA when having to speak in online tutorials, as much as 18% claim not to have any strategy to overcome it. The most used strategy is to prepare for the spoken interaction in advance by writing down the response (43%) – that is, to avoid speaking spontaneously. Many students claim to apply self-talk to regulate their emotions and overcome their apprehension. Around 20% choose not to participate by clicking the “Step Away” button (in Adobe Connect) or saying that their microphone is not working.

Figure 3.

“Which of the following strategies do you use if you have to speak in tutorials? Please choose all that apply.” (multiple choices; percentage of students, 307 responses).

The strategies that emerged from the qualitative data are in line with the quantitative findings and can be grouped into six broader topics.

3.2.5 Preparation

A way to alleviate FLSA in the online tutorials is to prepare for attendance; some of these overlap with the “study tips” described in Section 3.2.2. Quantitative data show that while one third of the participants of this study do not prepare in advance (Figure 4), over half claim to practise out loud what they think might come up during the online session. It is interesting to note that very few students seek support from their peers (8%). On the one hand, students would like to practise with people who are not “better” – that is, of more or less their level. Despite this, they usually do not see each other as a resource for practice. On the other hand, they consider only native speakers are worth practising with: “I speak to two native German speakers each week”. And although they expect help from their tutors, surprisingly only 3% actually turn to their tutors.

Figure 4.

“How do you prepare for speaking in tutorials? Please choose all that apply.” (multiple choices, 305 responses).

Others revise the materials in advance, if the tutor sends it to them, but would not ask for these; watching recorded tutorials is also a preferred mode of preparation (Section 3.2.2). Students are trying to find the strategy that works best for them. Some take private one-to-one classes or participate in a conversational exchange, but for many, the solution is not to attend tutorials: “I have not been attending live tutorials”; “I haven’t done a tutorial” (see also Section 3.2.6).

3.3 What are students’ speaking practices in asynchronous online language learning?

Finally, to respond to RQ3, asynchronous speaking practices have been analysed. While preparation practices also include asynchronous speaking, in this part we focus on how students work with the module materials and how they prepare for spoken assignments.

3.3.1 Strategies for module work

The proportion of students who claim not to engage with the listening and speaking activities at all during their module work is much lower than those who do not attend synchronous tutorials (Figure 5). The majority prefer working on speech perception by replaying the audio or using the transcript; fewer students are keen on practising speech production (speaking out loud; recording themselves).

Figure 5.

“Which of the following strategies do you use for your general module work? Please choose all that apply.” (multiple choices, 307 responses).

Teaching materials are designed so that students can practise speaking autonomously. However, one third of our respondents rarely or never do the recording activities (Figure 6) and, out of those who do, one third rarely or never actually listen to their recording (Figure 7).

Figure 6.

“I do the recording activities with a voice recording tool” (single choice, 307 responses).

Figure 7.

“When I do the recording activities, I listen to my voice” (single choice, 284 responses).

As for their feelings regarding the recording activities, about one third feel too anxious to record their answers (Figure 8) and about the same proportion feel confident not only to record but also to share their recordings with their fellow learners (Figure 9).

Figure 8.

“Rank the following statement about your speaking practice: When working through module activities online I am anxious about recording answers” (single choice, 306 responses).

Figure 9.

“Rank the following statement about your speaking practice: I am confident to share my recordings in the student forum” (single choice, 307 responses).

Independently of whether students do the asynchronous speaking activities or not, they dislike listening to their own voice (Figure 10).

Figure 10.

“Rank the following statement about your speaking practice: I like listening to my voice” (single choice, 307 responses).

4.1 Transitioning into breakout rooms: “A little virtual dance” begins

Researchers agree that the breakout room facility is invaluable, as it enhances collaborative learning (e.g. Bastos, Reference Bastos, Gómez Chova, López Martínez and Lees2022) and allows students to develop and apply their understanding of concepts that were explained in the main room (Tutton & Cohen, Reference Tutton and Cohen2024). In breakout rooms they can also raise questions or concerns with peers that they may have kept quiet about in the main online tutorial room (see Chandler, Reference Chandler2016; Tonsmann, Reference Tonsmann2014). Thus, breakout rooms can empower students. A novel finding of our research was that in the synchronous online context, one of the main triggers of FLSA are precisely breakout rooms.

Tutors can create a structured and mostly predictable learning space by managing conversational turns in the main online tutorial room and ensuring each student can get similar speaking time. This can create a false sense of security and lead students to expect that they do not need to initiate interactions themselves in online tutorials. Muñoz-Basols, Fuertes Gutiérrez, Strawbridge and Acosta Ortega (Reference Muñoz-Basols, Fuertes Gutiérrez, Strawbridge and Acosta Ortega2023), investigating the interaction in online coursework, concluded that student interaction and participation were overwhelmingly initiated by the instructors present. However, management of turn-taking in breakout rooms, with the tutor typically not present, becomes the joint responsibility of the breakout room participants, and assuming such agency can result in uneven, and anxiety-inducing, conversations where efficient learning might depend on others. The breakout room space can then appear unstructured, unpredictable, and out of the student’s control. Hilliard, Kear, Donelan and Heaney (Reference Hilliard, Kear, Donelan and Heaney2020) also confirm that the major cause of anxiety for students in an online collaborative project in a computing module were feelings of uncertainty. Navigating this space with other students can be a little virtual dance, where no one can predict who will lead or follow, which steps are next, and which direction the conversation may take.

Furthermore, students cannot hide as easily behind their computers as in the main room, which means they might not be able to apply the same mitigating strategies either. It might be more embarrassing in this space not to activate the camera or choose not to activate their microphones and in this way use technology as a way of opting out of contributing. The fear of group work may result in students leaving the tutorial or deciding not to engage in language activities in the breakout room. Topalov, Knežević and Halupka-Rešetar (Reference Topalov, Knežević and Halupka-Rešetar2023) concluded that the highest levels of anxiety were found where students took part in lessons by means of a camera and recommended that video communication should be reserved for small group discussions or pair work. Our data show that the use of cameras is not so straightforward in breakout rooms either, and students might prefer shielding even if it means feeling more isolated. It has been shown that viewing oneself during synchronous online sessions might increase appearance anxiety and lead to a decrease in learning (Tien, Imundo & Bjork, Reference Tien, Imundo and Bjork2023), while “Zoom fatigue” can also increase when cameras are used. Fauville et al. (Reference Fauville, Luo, Queiroz, Lee, Bailenson and Hancock2023) report that people online might feel that they are constantly watched so they must continuously produce nonverbal cues, besides having to constantly monitor others’ nonverbal cues. In their study, women are more likely to experience nonverbal overload than men. As generally the majority of language learners are female (72.3% of our respondents identified as female), the aversion towards the use of cameras found in this study might be worse than in other disciplines. Apart from this, for some learners, simply navigating the technological affordances of online tutorials can be anxiety-inducing, as they need to be able to interact with their peers and tutor through the use of cameras and microphones, as well as engaging through emoticons and writing in text chat or on whiteboards, and this can amplify FLSA.

Both the lack of direct physical contact with peers and limited visual cues result in lower levels of social cohesion. It has been noted that students may find breakout rooms far more demanding than working f2f (Peacock et al., Reference Peacock, Murray, Dean, Brown, Girdler and Mastrominico2012). In addition, comparing yourself to other students’ perceived levels of competence and confidence of speaking in the TL can invoke FLSA, and the work in small groups in breakout rooms can then distort these perceptions and accompanying emotions so much so that all the peers appear to be much better.

Hence, careful scaffolding is needed to enable a successful transition from main room to breakout room. Ensuring students know exactly what they are expected to do and how to go about it, including accompanying instructions in breakout rooms, can alleviate some fears during this transition. Douglas (Reference Douglas2023) warns that there is often a gap between what tutors regard as clear task setting and efficient breakout room management and how students perceive it. We found that the use of English by the tutor and with peers might be helpful in the transition process. Despite the widely held belief that in the foreign language classroom the TL should be used exclusively or almost exclusively, research on the applications of pedagogical translanguaging that challenges this starts to emerge (e.g. Adinolfi & Astruc, Reference Adinolfi and Astruc2017). The online context requires more explicit instructions and more scaffolding in student–student interactions. Translanguaging might be an efficient means to prevent communication breakdowns and achieve successful meaning-making when the adequate vocabulary is not available. This can involve the tutor explaining the task in English, and students asking for clarification and confirming their understanding in English. Or it can be just slipping in the odd word in English to ensure clarity or to keep the communication going. Our observations corroborate Reference Back, Han and WengBack, Han and Weng’s (2020) assertion that translanguaging can prove useful for effectively managing participants’ emotions in situations involving homogeneous groups. It can also be used to build rapport with peers and thus increase social support, which is important for anxious learners (Tutton & Cohen, Reference Tutton and Cohen2024). Interestingly, the social aspect of translanguaging was only marginally mentioned by the participants of this study.

Allowing some time for silent preparation before students go into breakout rooms might also help effective small group work. Yamagata-Lynch (Reference Yamagata-Lynch2014) points out that tutors should consider carefully the allocation of students to groups and providing a structured framework for the time spent in breakout rooms. In the context of the OU, tutors will not know all the students attending their sessions, and this will impact their ability to devise appropriate groupings for breakout rooms.

4.2 Vicarious learning

Although it is well attested in the literature that students need a learning community to thrive (e.g. Boling, Hough, Krinsky, Saleem & Stevens, Reference Boling, Hough, Krinsky, Saleem and Stevens2012), and that social connections might help to mitigate anxiety (e.g. McLean, Asnaani, Litz & Hofmann, Reference McLean, Asnaani, Litz and Hofmann2011), in the present research participants sought support from their peers only to a very limited extent. This is probably a direct result of the settings this research was conducted in, namely, that tutorial attendance is optional, so students might end up with a different tutor and different group of students every time, thus it is harder to build relationships and develop a learning community. Since a widely used preparation technique was to watch recorded tutorials, it might be a way to develop the feeling of belonging to the learning community.

While the results regarding the benefits of recorded university lectures are somewhat contradictory (see Pleines, Reference Pleines2020, for a review), Pleines (Reference Pleines2020) shows that, in contrast to the f2f classroom, in online and distance learning environments students who attend more live online sessions also access more recordings. She also found a significant positive correlation between the use of recordings and assessment grades, which is due to deeper engagement with the course. We conclude that vicarious learning is an effective FLSA mitigating strategy for our learners and there might be various reasons behind it. Recordings serve as a language practice resource to consolidate knowledge that students can use when they want to. They allow focus on linguistic forms and interactions in a low-stakes context. Literature shows that engagement with a variety of perspectives – in this case, offered by different tutors and their peers – can lead to deeper learning (Mayes, Reference Mayes2015). It also allows students to see that their peers have similar concerns and that they are not alone with their worries, which might reduce their feelings of isolation. The importance of role models in language learning has been long acknowledged (e.g. Dörnyei & Kubanyiova, Reference Dörnyei and Kubanyiova2013), which vicarious learning can provide. Our participants are used to learning alone and often prefer this mode of study. Still, recordings may help them position themselves within the group and strengthen their belief in accomplishing similar tasks. As Mills (Reference Mills, Mercer and Williams2014) notes, seeing the accomplishments of peers can boost self-efficacy beliefs. Vicarious learning and certain avoidance strategies are the two sides of the same coin. Similar to watching recorded tutorials, clicking the “Step Away” button to indicate to other participants that you are not currently at your computer or unable to participate means opting out of contributing actively, but still allows students to be present and listen to the tutorial – that is to say, stay engaged.