Responding to COVID-19 in EAP Contexts: A Comparison of Courses at Four Sino-Foreign Universities

The ongoing global COVID-19 pandemic has had an unprecedented impact upon education across multiple sectors, fields and disciplines. Campus closures, strict self-isolation and physical distancing measures require educators to rethink all aspects of course delivery. In response, many institutions have transformed face-to-face learning to online delivery at extremely short notice. Furthermore, for institutions delivering transnational English as a Medium of Instruction (EMI) education, the pandemic poses significant challenges for providing effective English for Academic Purposes (EAP) support. This paper aims to contribute towards the ongoing discussion related to COVID-19’s impact on English language teaching. Specifically, the paper presents a reflective overview of how five individual EAP courses across four Sino-foreign universities based in China, and at the forefront of the global COVID-19 pandemic, responded to this crisis. To provide a variety of contingency models, EAP courses at two Sino-US and two Sino-UK universities at different stages of development are discussed. is especially important for students engaged in transnational programmes, where developing academic language skills is only one of many ways that students must adapt to new academic expectations. encourages Chinese higher education English language instruction to learners’ cross-cultural communicative skills and international Quality Assurance Agency Higher Education’s are 1,240 Sino-foreign transnational co-operations with 22% and 21% of all Sino-foreign co-operations being with UK and US universities respectively (QAA, 2017). In relation to the ongoing COVID-19 pandemic, this means that successful Through reflection and comparison, areas of good practice and the key challenges faced when transforming EAP courses from face-to-face to online delivery are discussed. All five EAP courses are compared using interaction, learner autonomy, feedback, and leadership and institutional support as key themes. The paper also presents practical implications drawn from the shared experiences of EAP practitioners from all four Sino-foreign universities. Ultimately, the paper aims to share the lessons learned by EAP practitioners to support those who may be required to make similar adjustments to their course delivery both within and beyond EAP teaching contexts. as a second or

transition to online delivery and EAP support at China-based Sino-foreign partnerships could support the UK and US foreign partner universities during such uncertain times. Fortunately, as the COVID-19 pandemic has developed at varying rates across the globe, lessons can be learned from institutions which had to adapt first. For example, at the end of January 2020, 30 million learners across 300 tertiary level institutions in China had to pursue their education online (Lau, et al., 2020). These Chinese institutions adapted at extremely short notice without prior guidance, insight or specific models of good practice. Strategies adopted by Chinese educational institutions during this emergency transition can now be used as models for the other 176 countries facing similar ongoing educational disruption (UNESCO, 2020).
This paper aims to contribute towards the COVID-19 discussion and offer support to institutions currently transforming their education from face-to-face to online delivery. Specifically, insights from five individual EAP courses taught at four different Sino-foreign universities are presented. These Sinoforeign universities were at the forefront of the COVID-19 global transition to online education. For example, as the virus hit during the peak Chinese New Year period, tutors, students and staff were displaced across China and throughout the world. The paper presents areas of good practice as well as the key challenges faced when transforming EAP courses to online delivery, making comparisons across the sampled EAP courses. It is hoped that the insights presented will provide useful input and support for readers wishing to optimise their own online instruction especially during such unprecedented and challenging times for educators.

Teaching contexts
This study presents and compares individual accounts of praxis from six EAP tutors, across five EAP courses at four Sino-foreign universities. Specifically, the accounts focus on the unprecedented and rapid transition from face-to-face to online instruction. To provide a range of perspectives, the Sinoforeign universities are all at various stages of development and include two Sino-US and two Sino-UK universities. Additionally, although the focus of this paper is on the individual authors' courses and teaching, EAP courses from programmes of different sizes, at undergraduate and graduate level, working towards varied linguistic and academic goals are presented. Therefore, the approaches adopted and insights presented in this paper aim to provide a range of strategies for readers' consideration. Table 1 provides specific contextual information of the five EAP courses taught by each author.

Data collection and analysis
This study adopts a descriptive case study approach (Duff, 2008;Stake, 1995;Yin, 1994) that compares six sub-cases from the recent online courses of EAP practitioners (the authors) from across four different Sino-foreign joint-venture universities. Each of these teachers, at short notice, adapted their courses from face-to-face to online instruction. At the end of each course, the six EAP tutors wrote retrospective descriptive accounts of their teaching experiences, based on the following four categories deemed in advance to be particularly salient by the lead-author: -Interaction -Learner autonomy -Tutor, peer and self-feedback -Leadership and institutional support The authors then synthesised the descriptive accounts and drew out possible insights. As all authors were involved in the production, synthesis and comparison of the descriptive accounts, they were able to confirm the accuracy of items attributed to their individual EAP course teaching. The accuracy and appropriate interpretation of comparisons were further enhanced via follow-up online discussion and clarification amongst all the authors. The goal was to present a credible account of this unique transition to online teaching in order for other practitioners to identify practices and relevant details of the six cases that might profitably be transferred to their own contexts.

Interaction
The EAP programmes compared in this paper prioritise interactive and communicative, student-centered, pedagogies. These pedagogies increase interaction in the target language and promote successful Second Language Acquisition (SLA). For example, Swain's (1995) Output Hypothesis suggests that learners who demonstrate increased second language output will be better able to identify gaps in their own language development. Additionally, Long's Interaction Hypothesis stresses the importance of communication and negotiation for meaning for successful language learning (Long, 2015, p.52). Therefore, the authors design and teach the previously introduced EAP classes with second language interaction as a central focus. However, the sudden shift to online teaching and in particular asynchronous elements of online delivery present significant challenges to providing interactive and communicative EAP lessons. For example, although asynchronous online discussions can provide time-independent access to an ongoing discussion, limited student participation is a common problem (Cheung & Hew, 2004;Hewitt, 2005).
Despite the relevance of Swain's (1995) and Long's (2015) hypotheses for general SLA, they perhaps fail to take into account the affordances of online learning. In contrast, Siemens' (2005) Connectivism theory combines learning theory, social structures and technology to put forward a learning theory for the digital age. Siemens (2005) argues that when traditional learning theories are viewed through the lens of technology, they are no longer adequate. This is because the internet and online technologies expose learners to a rich array of information resources that they subsequently use to develop information networks beyond what is taught in traditional classrooms (Dunaway, 2011). Connectivism posits that learning occurs when learners make connections between concepts, ideas and perspectives accessed via internet technologies. As such, for online English language teaching contexts the concept of Connectivism should be considered alongside more traditional SLA theories (e.g. Long, 2015;Swain, 1995) to achieve effective learning.
Perhaps the biggest challenges to achieving successful interactive online delivery were faced by authors A, D, E, and F as their lessons were delivered asynchronously. However, authors A, D and E successfully used weekly asynchronous online discussion forums to encourage interaction between students and tutors. Managed through the universities' learning management systems (LMS), Sakai and Moodle, such asynchronous discussions allowed interaction and provided valuable feedback of students' engagement and achievement of learning outcomes. Generally, student engagement on discussion forums was high. Authors believe that this was because of clear instructions, tutor encouragement, tasks that involved commenting on other students' posts and a genuine desire to interact on the part of students. Additionally, tutors across all three asynchronous contexts sent regular emails and created personalised videos to support students. Furthermore, authors A and F supplemented asynchronous class input with frequent synchronous office hours, where the tutor was online for students to casually "drop-in" as well as scheduled individual or group tutor meetings. These additional channels provided essential interactive support for students to clarify lesson input and receive advice on assignments as evidenced in student feedback.
"Every week he would at least once encourage us to meet him during his office hour. I get the safety that he is always there to help." In contrast, authors D and E did not incorporate live sessions into their course, although optional live tutor-facilitated discussion sessions were offered to students midway through the semester using Zoom. Student feedback on these sessions was positive but uptake was limited and seemed to decline as the majority of students returned to campus towards the end of the semester.
Furthermore, to encourage interaction amongst students, authors A, B, C, and F developed student learning teams. For example, authors A and B adopted collaborative student writing groups of three or four students for the whole course. This created an additional support network other than the tutor. Students were required to meet frequently in their writing groups on Zoom, share screens and discuss their ongoing assignments using English, the target language. Importantly, students recorded and sent their meetings to the tutor for feedback, which motivated learners to remain on topic and use the target language. Further strategies for achieving asynchronous interaction involved the creation of personalised interactive screencasting video voiceovers to supplement class PPTs.
"The study group is great! It helps me gain deeper understanding of our readings!" "Group discussions promote communication and the sense of engagement, which is very useful." Authors B and C both adopted a mixed synchronous and asynchronous approach to online delivery and therefore had more scope to achieve successful interaction. Both authors adopted Zoom and despite students' geographical dispersion, poor internet connectivity did not hinder the online synchronous lessons. Generally, both tutors found that student-to-student interaction in discussions, both in whole class mode, and in Zoom breakout rooms, was lively. Noticeably missing was movement around the classroom and tasks that depend on such movement, such as gallery walk activities where students move around the classroom to explore and discuss different texts displayed in different locations (similar to an art gallery). Author C was able to mimic such activities using Zoom's ability to set up a string of Zoom breakout rooms simultaneously, permitting students to move from one to another. Similar to author A, both tutors successfully utilised the record function of Zoom. This enabled students to complete many tasks centered on spoken interaction and then submit these to a shared folder online for instructor review, and further peer review tasks. Another useful function of Zoom exploited by author C was the ability to turn the tutor's camera on and off throughout a synchronous lesson. This allowed the instructor to "disappear" at key moments, while still maintaining a presence and observing the lesson, akin to good monitoring practices where the instructor monitors from varying distances.
Like authors A, D and E's asynchronous context, authors B and C also adopted asynchronous discussion forums to supplement their synchronous delivery. Author B opted for weekly written discussion forum posts using both Sakai and Padlet, as this fitted the EAP course aims of developing students' academic writing skills, yet incorporated multimedia to make discussions more stimulating. However, author C used Voicethread for "threaded" voice discussions, in analogy with discussion boards. This linked to the oral production learning outcomes for that particular course. Additionally, in order to support learners and encourage further face-to-face interaction using the target language, author B scheduled additional optional synchronous Zoom meeting sessions. During these sessions, students were encouraged to meet their tutor informally as a class to hold discussions using media related to course topics (i.e. TED talks) over a cup of tea. Finally, authors A and B successfully used Zoom's virtual background feature for asynchronous and synchronous lessons. For example, using different locations that were personal to the tutors seemed to reduce students' negative feelings which hinder learning, or affective filter (Krashen, 1985), for self-isolating students. Students' final course reflection assignments and course evaluations praised this practice. For example: "I enjoyed so much the beginning background of every online session you prepared. The scenes helped a lot especially in today's global situation with house quarantine. They kept me happy and optimistic. Thanks for preparing these." "For every online session, he sets his background as different scenes he has been to, and takes one or two minutes introducing them to us. This is a minor point about this course, but it shows his enthusiasm and really makes me happy taking online sessions. As the world is experiencing something bad, his openings always make me optimistic."

Learner autonomy
The EAP programmes compared in this paper also aim to help students become more independent learners when making the challenging transition to university and later to graduate school. The teacherfronted, prescriptive and test-focused pedagogies experienced at the secondary-school level by many Chinese students as described by Hsu (2015) warrant this focus on developing learner autonomy, which is often closely linked with the concept of taking control of one's learning (Benson, 2001;Little, 1991). Moreover, within Sino-foreign universities, the transition can be even more difficult as learners must quickly adapt to the academic expectations of the foreign host university. Furthermore, the aforementioned student-centered, interactive pedagogies and increased expectations of student independence present additional challenges. A further complication is raised by the fact that learner autonomy may be culturally conditioned. For example, Cao (2011) highlights that students of East Asian Confucian heritage are unlikely to demonstrate autonomous learning styles due to cultural constraints. Therefore, as recommended by Borg and Alshumaimeri (2019), the authors design and teach their EAP courses in order to develop learner autonomy. This goal becomes even more important when moving to online instruction at short notice. For example, Schweinhorst (2007) notes the importance for learners to develop metalinguistic awareness, self-reflection, critical thinking, and experimentation while studying online.
The sudden and unplanned increased need for learner independence was a concern shared by all authors. This was particularly important, given that the EAP programmes described all aim to foster the very student autonomy that asynchronous portions of online delivery require. For example, although tutors utilise their "teaching voice" to scaffold and support, learners must be able to manage their own learning when studying foreign language via distance (White, 2014). Authors A and B initially noticed that many learners were unable to manage independent study workloads. For example, many students were displaced without learning resources and initially spent more time than was allocated on class tasks, failed to complete homework assignments on time and generally seemed overwhelmed with the sudden change and perceived increase in workload. For these two EAP courses, the shift to online delivery occurred mid-way through the semester after collaborative group projects had begun. Therefore, this may explain why such learners found the transition so hard. In response, both authors adopted consistent weekly checklists of class tasks, homework, asynchronous discussion forum tasks and assignment deadlines. Following positive student feedback, authors A and B later developed the checklists into student self-evaluation tasks to facilitate reflection and for learners to take ownership of their learning.
Similarly, author C successfully implemented weekly homework, task, and assignment logs. The university requirement was for students to spend 10-12 hours per week on coursework and tasks outside of class. Students made their own choices about how to fill time wisely in order to target the course learning outcomes. The logs allowed for later feedback and discussions with the instructor and peers on the effectiveness of the time spent and possible alternatives.

"…the instructor clearly shows the checklist of the whole week which is useful for me to schedule my time."
Authors D and E intentionally designed their EAP course with a framework for learner autonomy. Each week had clear, consistent deadlines highlighted on the EAP course Moodle page and in a weekly email from tutors. Moreover, authors D and E designed and sequenced tasks to make each activity meaningful. There was a clear routine, established early in the course and maintained throughout. This was a priority because like tutors across all four institutions, authors D and E anticipated that students would find it difficult to quickly adjust to online learning. However, students took to this remarkably well and after initial teething problems a clear majority of students were independently managing their work on the EAP course. This was supported by a student feedback survey in which just under 88% of respondents agreed or strongly agreed that the module was "well-organised" and had been "running smoothly". Students commented that: "I found (the module) most engaging this semester due to the clearly structured course arrangement and the video assessment which helps me to improve my oral speaking skills." "every week I receive a pdf guiding me what to do during the week, which makes me clearly understand the tasks and learning objectives" Author F teaches the lowest level of EAP students and was concerned how learners would manage without face-to-face classroom instruction. As such, the tutor built support and teamwork mechanisms into the asynchronous EAP course design. For example, students created WeChat groups of three or four to discuss academic challenges and collaborate on team-based coursework. Author F also staggered materials so they were only available for seven days to help learners manage their time. Furthermore, authors B, C, and F developed WeChat groups for each individual EAP class for learners to meet and discuss as a class with their tutor, and serve as a convenience function (for communicative reminders), and as a contingency communication platform. Following good scaffolding practice, author F used these WeChat support groups frequently in the early weeks of the new semester but less frequently as time progressed. On reflection, author F's EAP students excelled and the submission of coursework improved in terms of quality and punctuality. Students appreciated the relative freedom and apparent trust that asynchronous delivery gave them to manage their time.

Tutor, peer and self-feedback
The compared EAP courses all have carefully planned programmes of feedback from tutors and peers. For instance, as many first-year nonnative English-speaking undergraduate students are unfamiliar with the genre and style conventions of EAP communication, regular feedback is required. According to Schmidt's (1990) Noticing Hypothesis, feedback can also facilitate language learners' awareness of gaps between their current interlanguage and the target language. Furthermore, Hyland and Hyland (2006) highlight the beneficial role feedback plays in both second language writing skills development and learner motivation. Feedback from spoken interaction and negotiation for meaning between second language learners and tutors can also facilitate language learning (Long, 2015). As such, the authors design their EAP courses to include a balance of feedback mechanisms to deal with classroom interactions, oral presentations and written assignments. However, the sudden shift to online teaching required tutors to modify feedback mechanisms. Although digital technologies have been shown to support written feedback (Bakla, 2020), there is limited consensus on which approaches are best.
As all six authors were teaching students across different levels and disciplines, and were working towards different academic goals, the EAP courses adopted various online feedback mechanisms. However, for authors A, D, E, and F, asynchronous delivery meant that tutors could not gauge students' comprehension of class input in real time. As such, it is important to record asynchronous class input videos in an interactive way and regularly ask concept-and task instruction-checking questions that are then clarified. In contrast, with synchronous classes, tutors are able to evaluate students' comprehension and provide live feedback. For example, author B used Zoom's participants voting and polling features to frequently ask whole class comprehension questions. This strategy motivated students to remain focused throughout synchronous lessons in preparation to answer questions, much like successful face-to-face classroom management.
In terms of providing feedback on students' EAP assignments, online modes of delivery provided multiple possibilities. For example, authors A and B increased formative EAP assessment to provide more opportunities for learners to receive and act upon feedback. Both tutors successfully utilised Turnitin to ensure academic integrity of online written submissions and to supplement in-text comments and rubric scales with personalised voice feedback as recommended by Keane et al. (2018). Such feedback mechanisms were well received by students and appeared to facilitate uptake of feedback. "Peers are one of the audiences of our articles, so their suggestions on content, logic and readability are very useful." However, author F observed through formal Turnitin and informal written academic submissions, increased incidences of plagiarism where students passed off other students' work as their own. Authors D and E noticed an increase in students copying large chunks of text directly from online sources either with or without correct citations in their oral presentations. Such increased incidents of plagiarism and failure to use appropriate academic citation and paraphrasing skills may be due to increased workloads and the previously discussed challenges associated with increased learner autonomy.
As author C's EAP course was delivered synchronously and focused on oral production skills, oral feedback mechanisms were preferred. As such, Zoom allowed the instructor to provide oral feedback to students and for students to provide feedback to one another in both synchronous sessions and during feedback tutorials. Voicethread added an additional channel for asynchronous feedback. Author C designed Voicethread assignments so that students received oral feedback on their performance and input into the creative process of carrying out their semester project. For example, twice across the semester, students recorded a three-minute self-evaluation using Voicethread and their instructor and classmates then built on the evaluation. Additionally, authors A, B, C and F all successfully used synchronous tutor office hours to meet with students and provide feedback on students' assessments and overall course progress. It was generally agreed amongst tutors that such online feedback channels equaled face-to-face student-tutor meetings, or perhaps even increased uptake due to reducing the strength of the affective filter (Krashen, 1985). However, further research is required to confirm this comparison.
Regular feedback was a key part of authors D and E's asynchronous EAP course design. However, the tutors altered the initial course plan to give video feedback on students' individual work, as the videos were too time-consuming. Instead, to maintain a sense of familiarity they made videos giving whole class feedback. Students praised these feedback mechanisms in their end of semester module evaluation student survey. Six qualitative comments explicitly mentioned feedback, and five of these were positive, including: "we are provided with some specific skills and weekly feedback to help us correct our faults and improve step by step" Similar to authors A, B and C, peer and self-feedback were frequently used in authors D and E's courses. Each week, a group feedback task required students to watch videos of their groupmates and comment on their content in a reply video. Additionally, for lecture response tasks, students provided feedback on a partner in an online forum. Authors D and E made a video modelling how to give effective peer feedback following '4 Cs' of concision, kindness, clarity and constructiveness. The quality of peer feedback in these forums was striking to the authors, perhaps in part because of the students' frequent exposure to similar tutor feedback and the guidance given in the videos. Finally, reflection formed a key component of authors A, B, C, D and E's EAP courses. For example, author C explicitly taught reflection using Gibbs' (1988) 'model of reflection'. Authors D and E asked students to reflect weekly on their performance and the feedback they had received from the previous task and incorporated a final reflective written task based on Rolfe et al. (2001). This encouraged students to attend to their feedback, flagged instances where feedback had not been received and promoted use of feedback to guide later work, or feedforward.

Leadership and institutional support
Before addressing the key practical implications of the insights presented in this paper, it makes sense to reflect collectively upon the shared experiences of authors across all five programmes in terms of leadership and institutional support. Firstly, there are several areas of good practice from an institutional level that should be commended and where possible adopted by senior administrators to facilitate any move to online instruction. For example, when institutions acted early and communicated clear and wellconceived plans of action to tutors and students, it reduced anxiety and allowed more time for tutors to plan materials and adjust EAP curricula. Furthermore, the Sino-foreign universities discussed have many international teachers. Therefore, as the COVID-19 pandemic spread during the Chinese New Year holiday, many tutors ended up displaced overseas where they were on vacation without access to key teaching resources and personal belongings. Some institutions supported their staff financially by funding accommodation, travel and the purchase of equipment such as laptops and microphones. This further reduced anxiety and allowed tutors to focus on teaching.
Linking to such anxiety, students and tutors had to endure uncertainty, displacement and social isolation all while teaching and studying in an unfamiliar environment. As such, institutions who offered counselling and emotional support to students, faculty and staff should be commended. Examples of this observed by the authors range from anonymous formal online counseling services; live one-toone counselling sessions via video call, one-on-one tutor meetings with senior university leadership and departmental directors; regular university-wide faculty meetings, departmental meetings and course-level meetings; tutors in larger course teams being assigned a group coordinator; and online social support meetings amongst colleagues. Finally, in order to gain an accurate picture of student and tutor displacement and to manage synchronous schedules across multiple time zones, frequent and comprehensive surveys were conducted to keep track of university stakeholders' changing personal circumstances.
In terms of pedagogical, course design and technical support received from institutions there are also several areas of good practice to report. Some universities created and distributed detailed and researchinformed guidelines for online teaching best practice. Such support and guidance proved extremely helpful for tutors initially, and was a useful point of reference throughout the semester. Furthermore, universities with designated instructional technology services were able to provide extremely useful advice, recommendations and support related to educational technologies. For example, the authors benefited from: • online depositories of university-sponsored educational technology tools, user guidelines, and suggestions. • online discussion forums for teachers to share areas of good practice, ask questions about educational technology tools and receive answers from specialists. • institutional subscriptions to an increased number of educational technology licenses. • online university-wide conferences where teachers shared ideas and areas of good practice. • provisions for tutors to participate in online courses provided by the foreign partner university and regular online workshops focusing on various educational technology tools. Additionally, institutional support for tutors engaging in continued professional development courses related to online teaching pedagogy proved particularly helpful. Finally, for larger EAP teams, universities implemented management and coordination schemes in which members of small management teams acted as points of contact supporting a larger group of tutors. Moreover, materials designers created common resources for all students which they hosted on an LMS. This reduced duplication of tutor workloads during times of increased anxiety and ensured standardisation of EAP course input.
However, as expected given the unprecedented nature of this emergency, institutional support may be strengthened by consideration of several further areas. For example, although the universities did their best to keep all stakeholders informed of the COVID-19 developments and the ongoing plan for teaching, some were more explicit than others. This created uncertainty for tutors receiving more ambiguous communications. In times of crisis it is important to reduce anxiety and provide clarity and stability where possible so that teachers can focus on teaching. However, as the situation was rapidly developing and as there was no model to follow, it is understandable that universities may not have had all the answers. Additionally, despite university-wide policy and guidelines, some non-EAP courses adopted start dates and online delivery methods that were not in line with the original university policy. This made scheduling more challenging and meant that learners had to use different online tools to access classes. This caused frustration for some learners who were already finding the increased learner autonomy challenging. Finally, unlike a bustling academic campus, the online teaching environment can feel isolating for tutors. As such, institutions could have increased the frequency of departmental meetings to increase EAP specific synergy and support. However, tutors were spread out across multiple time zones and were extremely busy and may not have wanted to attend such meetings. Perhaps, forming tutor groups based on similar time zones could have provided a workable solution to this problem.

Implications for Future Practice
The final section of this paper will summarise the key reflections from all five EAP courses and present implications for future practice. Such implications should be practically applicable for readers adapting their teaching from face-to-face to online delivery in both EAP contexts and beyond. Firstly, tutors should prepare for the increased need for learners to study independently and autonomously by designing their online courses with support mechanisms in place to help students overcome such challenges. For example, they could achieve this by: • using consistent deadlines and task patterns.
• creating familiarity with learning tools and online LMS systems through, for example, step-bystep written instructions or video guides. • ensuring contingency submission mechanisms are available and clearly explained. • providing clear guidelines in terms of deadlines, weekly tasks, workloads and expected time commitments. • using specific weekly checklists that double up as reflective self-evaluation tasks to increase learner agency. Enabling and encouraging students to take ownership of their time management and to evaluate their performance also has the potential to increase learner autonomy. This links to the goals of many EAP courses and will ultimately help students succeed when studying in an online environment.
As demonstrated within this paper, no one-size-fits-all approach to successful online transition is possible and an openness to experimentation and adaptation is beneficial. The decisions made by programme administrators and tutors should take into consideration multiple factors. For instance, one must strongly consider the course learning objectives and then decide which strategy is more likely to produce outcomes linked to these objectives. For example, for EAP courses with a focus on academic writing, written asynchronous LMS discussion forums would seem wholly appropriate to reinforce critical academic reading and writing skills. However, for EAP courses focusing more on oral production and academic discussions skills then Voicethread discussion and video responses are more appropriate to reinforce listening and speaking skills. Furthermore, the geographical dispersions and associated global time zones of international tutors and students as well as participants' levels of internet access should also be strongly considered before committing to synchronous or asynchronous delivery. For courses adopting asynchronous delivery, it is important to note that it significantly decreases interaction. Therefore, it is vital to plan and include regular opportunities for student-student and student-tutor interaction. This can be achieved by: • offering live individual or group office hours, conferences or informal online meetings.
• scheduling regular asynchronous discussion forum tasks.
• utilising the record function on video communication software for students to meet and collaborate on projects. • creating small student support groups who work together throughout the semester. • ensuring asynchronous PPT voiceovers are as personalised and interactive as possible. Furthermore, given the online nature of such EAP instruction, tutors need to be aware of and try to reduce instances of plagiarism such as students passing off classmates' work as their own or directly copying large chunks of text from online sources without proper citation. Increasing formative assessment and providing increased feedback on assignment drafts can help students feel supported and increase tutors' awareness of students' individual academic styles. Additionally, increasing peer and selfevaluation of assignment drafts can increase the quality of students' academic work while promoting learner autonomy. For EAP instructors working with written assignments, Turnitin can be a deterrent and a useful tool to avoid plagiarism and raise student awareness of how to effectively paraphrase, rather than directly copying large chunks of text from a source. Additionally, the PeerMark feature is an effective tool for facilitating online peer feedback tasks. However, for EAP courses focusing on oral production skills, how to effectively deter students from reciting long chunks of source texts rather than putting them into their own words is a challenge that needs further consideration. To summarise, online teaching provides tutors with multiple opportunities for providing detailed, clear and personalised feedback, but potentially increases instances of plagiarism.
Finally, it is important to share examples of good practice and to reflect upon the challenges faced at course, programme and institutional level both internally and externally when adopting new modes of educational delivery. Some universities collected individual EAP tutors' experiences adapting to online teaching and shared these at different stages of the semester and at different levels. For example, at the course level, EAP tutors reported individual successes and challenges that were shared amongst EAP course teams (See Appendix 1). These results were also summarised and shared at departmental level (See Appendix 2), which was helpful as the insights all related to language instruction. Furthermore, to receive increased perspectives beyond discipline specific tutor feedback, all university tutors completed formal feedback surveys to share their experiences with colleagues across the university. It is pleasing to see that such useful insights have recently been shared with foreign partner university campuses to help support their later transition to online teaching due to the COVID-19 pandemic.

Conclusion
In conclusion, this paper has presented and compared the individual experiences of tutors from five EAP courses taught across four Sino-foreign universities at the forefront of the COVID-19 global transition to online education. Such insights highlighted areas of good practice and the key challenges faced by tutors. The cross-institutional comparison of EAP courses was discussed in relation to the themes of interaction; learner autonomy; tutor, peer and self-feedback; and leadership and institutional support. Key practical implications linked to these themes were then discussed for readers' consideration. It is hoped that this paper can contribute towards the discussion of how the ongoing COVID-19 pandemic is affecting English language teaching globally. The insights and implications presented should provide further support and input for readers currently teaching English online or for those considering adapting to online delivery. Finally, by sharing the experiences of EAP tutors across a variety of courses and programmes this paper aims to encourage others to engage in the discussion and offer support during such unprecedented and challenging times for English language educators.

Appendices
Appendix 1: Example EAP Tutor Insights Shared with Departmental Colleagues in the Early Stages of Online Transition

EAP Tutor A * What is going well
• Voiceover PPTs using Zoom for class content with class activities planned and task instructions clearly given. I then ask students to pause the video for the time that I suggest to complete the task. Students then stop their video, go away, complete the task and return. I then provide some model answers or points of discussion. • A brief welcome video before each session with my face/video on. This is an informal greeting to make the sessions feel more personable. • I try to include comments on the online lessons to make the lessons feel more personable. I regularly mention the names of students who have done a good job in their homework, relate points to previous lessons etc. I try and make the lessons specific to my students as opposed to a general online MOOC that anybody could take. • Some light comments help. I often ask students to take a break and put the kettle on to make some tea etc. • Providing a weekly checklist of lessons, tasks, homework, online discussions, assignments and deadlines. Students have to complete the checklist when they finish a task with Y (yes) L (if they completed work late) or N (if they failed to complete a task). This weekly checklist is submitted with their weekly homework in an attempt for learners to take ownership over their work and time management. • Weekly discussion posts on Sakai. I set a task and students must make an individual post and respond to at least 2 other classmates. I will regularly post responses to engage students and encourage further discussion. • 2 hours of live 'drop-in' office hours each week at a time when all students are available on their schedule. It's important that this time is when all students are available. • For student group projects, I have asked them to hold team Zoom meetings to discuss and work on the project. Students must record their meeting and send it to me with the audio transcript. This ensures that they work on the project, lets me see that everyone is involved and on track and gives me a sense of where they are in the process. • For group projects and team meetings, to give students a focus, I ask them to come up with 5 specific questions that they want to ask me about their group writing project by the end of their meeting. These 5 questions are submitted in writing with their video. • Regular written announcements via Sakai with clear instructions of tasks and deadlines. Note the time zone, day, date, and time to be specific. • Using Sakai 'Assignments' area for students to submit 1 set of homework each week, this is easier than organizing 2 submissions each week when students might be on different asynchronous time schedules for completing classes/homework. • I physically upload my Zoom video file as some students are unable to stream the Zoom cloud web link. • Encouraging students to post their general queries or questions about assignments on Sakai. The 'General Discussion' forum works well as it saves me repeating myself and allows other students to benefit from the questions/answers posted.

* What isn't going well
• I'm unaware of what else my students are doing and when. I don't know about their other classes, assignments/tasks, deadlines etc. so worry that they may be overworked. • Some students are very engaged and clearly managing whereas others are very quiet and submit tasks late. I reach out to these students to check they are OK and have introduced a weekly checklist in response to this. • It is a challenge to ensure students complete the classes by the suggested deadline. As my lessons are asynchronous, I suggest learners can complete lessons before their scheduled lesson time but not after. However, it is clear that still some students are not managing their time. This then affects the quality of online discussion tasks. • Students at the start were unable to follow simple instructions despite their clarity and found it hard to manage where and when to submit different tasks. Regular reminders help and hopefully the checklist will help. I start my online lessons with a logistics/reminders section. • I find the online teaching much more time consuming in terms of planning, executing and managing than face-to-face teaching. • Balancing workloads/timing is a challenge due to the online nature of instruction. Emails, homework submissions and Sakai discussion forum posts will come in at all times throughout the day. I think I need to be careful and ensure I set specific times to deal with each element and try to fix a regular working schedule. At the moment the boundaries between work hours and leisure are very blurred. • Although I have set tasks in the lessons that students are asked to submit, I do not have the same grasp in an asynchronous delivery of which students are successfully completing tasks and how long it is taking students. When teaching languages, in class student feedback and task performance is crucial and I worry that via online delivery I could easily miss a student who is struggling or who has not grasped a task correctly. Similarly, a student could race through the video without spending the necessary time actually completing a task and I would not know.

EAP Tutor B
* I conducted a short Qualtrics survey at the start of this week and the feedback I received from my students was largely centered on the following: • I have been providing a weekly checklist to help students know what they need to complete and the overwhelming feedback was that this was helpful. • Students liked the step-by-step approach to input (well-structured and clearly signposted). • Course workloads in general (not necessarily EAP) are higher since moving to online delivery and students need time and flexibility to (a) adjust and (b) keep up. • Students also seem to be responding to regular reminders, increased feedback on drafts and the overall flexibility I've tried to have for communication and support.
• Group conferences (for group work) on Zoom worked well and I'll be encouraging more Zoom meetings at the start of Session 4 (to set a precedent). • Office hours: I have had some students 'drop-in' on Zoom, but largely, students have been preferring to make appointments at times convenient for them which has worked well this session, (but needs more consideration with the updated synchronous input preferences for session 4). * From my personal experience: • A tip I got from a colleague was to ask students to ask specific questions about their assignments in the margin of a paper and then I respond to those during tutor feedback (as opposed to the whole paper). This feedback method has proved productive. • After peer review/ feedback on drafts, I asked students to highlight in their next draft (submitted to me) two changes they've made as a result of feedback -this gave them ownership on their improvements and indicated to me that they had: (1) completed the peer review process (2) learnt something useful form the experience. • Some students are still reticent when it comes to Zoom -it depends on the individual. As the shift to online learning was so sudden, I wanted to be as supportive as possible by providing as many channels for communication as possible (to appeal to all needs/ support everyone). This has been very demanding and (although very popular with students based on feedback) I will streamline channels of communication in the future (now that we all have an idea of what works and what does not). • Regarding forums, they can be very useful, and I assigned tasks in staggered ways so that everyone had to post by deadline 'A' and then respond by deadline 'B' (to ensure that there were plenty of threads to respond to / discuss, and it wasn't all done in the 11 th hour!). Yet, the quality of responses wasn't great (and after discussions with colleagues, it is hard to assess the quality of discussion forum posts -especially if they are only weighted for participation). The course weighting assigned to such tasks will be given more consideration as we move into session 4 (and we have more flexibility with setting the syllabus/ designing assignments that better suit online delivery).

Appendix 2: Faculty Feedback Summary Sent to All Faculty across the Department
Lessons Learned -Highlights 1) Things that worked well a) Require meetings with individual students/small groups of students instead of only holding dropin office hours on Zoom b) Have a weekly study guide, to-do checklist, or a compiled folder that helps students have a clear overview of the weekly deadlines, routines, announcements, assignments c) Clear expectations of deadlines, materials, and assignments d) Give students time to adapt and make mistakes e) Use a variety of options to engage all different types of learners. Some students thrive with asynchronous classes, written interaction and participation while others have trouble handling the freedom and distance f) Frequent communication to keep a sense of community g) Take a more student-centered approach in synchronous online classes. For example, ask students to lead discussion or report on progress h) Encourage student participation in online sessions with both video and audio on i) Make online instruction more personable j) Give peer review and peer feedback assignments k) Open-book exams that focus on skills and performance more than memorization l) Technology tools 2) Things that didn't work very well and/or were challenging a) Drop-in Zoom office hours generally have low attendance rate b) Keeping the balance of "required" and "flexible" tasks is challenging. Students' access to resources and individual learning motivations vary. c) Hard to manage differences in student learning motivations and sustain student motivation d) Time management and balancing workload -for both students and faculty e) Maintaining/ensuring academic integrity f) Allowing multiple channels for students to submit homework g) Hard to monitor how, when, and how well students are understanding and completing tasks h) Technical issues i. Internet connection ii.
Limited access to websites due to internet censorship (both in China and abroad) 3) Things that worked but take a lot of time -we may need to cut back or improve efficiency a) Recording lectures b) Giving students individual feedback c) Making videos on VoiceThread d) Content creation e) Sakai "Forum" -time-consuming, and hard to assess f) PPT voiceovers g) Grading h) Warpwire: takes too long to upload videos i) Planning, executing and managing online sessions i. Have a welcome video before each session or welcome students with your video on ii.
Include personalized comments, mention student names iii.
Break time during sessions iv. Zoom (useful functions include "Breakout rooms", drop-ins) v.
Voice-over PPT vii.
WeChat (useful functions include WeChat groups for class communication, and using voice message for pronunciation practice in Chinese class) viii.
Suggestion Box: For collecting questions, feedback and suggestions xi.
Otter (for generating transcripts)

4) Other suggestions:
a) Strengthen collaboration among instructors to reduce individual burdens (on content creation, etc.) b) Do fewer papers but more papers with multiple drafts so students get in the habit of taking in feedback c) Emphasize academic integrity d) Put limits on the amount of time and effort we put in. Yes, we want to give students good learning experiences even under these difficult conditions, but we also need to pace ourselves and avoid burn-out