Flipping a Virtual EFL Public Speaking Class Integrated With MOOCs During the COVID-19 Pandemic

This case study explored Chinese undergraduate EFL students’ attitudes to and perceptions of an online English public speaking course, which employs a virtual flipped classroom model and MOOCs during the COVID-19 pandemic outbreak. Since all classes were moved online, a previously flipped public speaking course integrated with MOOCs was converted into a virtual flipped classroom. All 25 participants of the study were undergraduate students in the science, technology, engineering, and mathematics (STEM) field. Zoom, Blackboard, and QQ instant messenger were platforms utilized in instruction. There were weekly two-hour Zoom meetings with learning activities using MOOCs on Blackboard. The study collected and corroborated results from multiple data sources, including surveys, focus group discussions, student presentation videos, and the instructor’s reflective teaching journals. Data was analyzed using Charmaz’s (2006) grounded theory. Survey results indicated that the 25 participants generally felt positive about the virtual learning environment. Students strategically adapted to all three digital platforms (Zoom, Blackboard, and QQ instant messenger), the MOOCs, and the flipped classroom model. They were engaged in exploring a variety of digital platforms, online learning resources, remotely collaborating with peers and interacting with the instructor. Incorporating MOOCs in a virtual flipped classroom allowed for application of theory into practice under the instructor’s supervision, which maximized the students’ speaking and learning opportunities. Recommendations for ELT practitioners and further research are also provided.


Introduction
Since early 2020, the COVID-19 pandemic has been spreading and disrupting formal education across the globe. As of March 24, 2020, 80% of the world's 1.37 billion students had been affected by school closures in 138 countries (UNESCO, 2020). To reduce the spread of COVID-19 through safe distancing, there has been increased reliance on online instruction. This case study describes English language teaching practices used at a university in Southern China in response to the sudden transition to online learning during the pandemic. In early February, 2020, due to COVID-19, the university administration announced that all 2020 Spring semester courses from February to June would initially be moved online. All instructors were required to teach online using Blackboard, an educational interface that connected students with the instructor, each other, and the learning materials. Video conferencing platforms such as Zoom and Tencent Meeting, and messaging apps such as QQ and WeChat were supplementary tools for communication. At first, instructors were required to prepare a 4-week online teaching plan. However, soon, the online course was extended to the entire 15-week semester.
This case study concerns a blended English language public speaking course using MOOCs and the flipped classroom model that was moved online. The course was an elective designed as a complement to the general English courses and English for Academic Purposes (EAP) offered to undergraduate students in a four-year science and engineering degree program. Most students who enroll in this course major in the field of science, technology, engineering, and mathematics (STEM) and have completed the requisite EAP course. Even though the students have given presentations in their general English and EAP courses, most of them have never had formal training in public speaking. This course focuses on the development of students' professional English communication skills. This case study explored students' learning experiences in the online version of the course, which may encourage further discussions on how to manage transitions to online teaching.
The rest of the paper comprises five sections. The next section 2 describes the pedagogical challenges that the instructor and students faced and the solutions to these challenges. This is followed by the methodology (section 3), discussion of findings (section 4), and the conclusions and implications (section 5).

Teaching Context
The abrupt transition to online teaching presented challenges to both the instructor and the students in the public speaking course described in this study. Before COVID-19, the course consisted of a two-hour face-to-face (f2f) class meeting each week for 15 weeks with 16 online modules covering 14 weekly topics, along with two extra modules for the midterm and final preparation on Blackboard. Students were expected to prepare for class each week by watching video lectures and completing assigned readings on Blackboard. The embedded recorded MOOC lectures were selected from a popular public speaking course on edX. Class meetings were devoted to group discussions, group presentations, peer evaluations, and instructor feedback. During the semester when the COVID-19 outbreak occurred, a total of 30 students who were new to the public speaking course enrolled in the two classes offered by the instructor, with each class comprising 15 students. Although the pre-pandemic version of this course had already contained online components, a sudden shift to completely online teaching still posed challenges in terms of implementation.
First, no one could predict when f2f teaching would resume, so the entire structure of the course was adjusted as needed. For example, students were previously expected to conduct group presentations in Week 4. However, since the university administration mandated that all courses should begin with a fourweek online teaching plan, the instructor tentatively moved the group presentations to Week 6. However, as Week 6 approached, a return to normalcy still looked unlikely, so the group presentations had to be conducted online.
Second, an unexpected problem occurred when Blackboard was inaccessible in Week 5. The instructor had to upload all teaching materials to QQ and cancel Week 5's discussion assignment following the lesson preview. Thanks to the file-sharing feature on QQ, Week 5's instruction was still effective, yet this issue highlighted the need for a contingency plan to overcome unexpected hurdles.

Zhuo Li and Xiaodi Zhou
Third, in order to hold students accountable for their learning, in Week 1's module on Blackboard, the instructor clarified her role as a facilitator and urged students to take an active role in learning. The students were also informed that the instructor could be reached via multiple modes of communication such as email, QQ, and each module's Q&A forum whenever they had questions. The descriptions of the roles of the teacher and students on Blackboard are shown below: The instructor's role: As the instructor of this course, I'm here to facilitate your learning. You can reach me by email, QQ, or each module's Q&A forum on Blackboard. I strongly recommend that you use the Q&A forum if you believe your peers may have the same question.
In doing so, my responses would help all others. Students' role: You need to be actively involved in your own learning. Please finish readings and view video lectures as required before each class meeting. Communication is vital to online learning. If you have any questions, please do not hesitate to contact me.
In addition, the primary challenge of transitioning to online teaching was the lack of substantial f2f interpersonal interaction. Even though Zoom could provide access to live-video communication, this alone could not compensate for the lack of in-person engagement. Still, the instructor sought to leverage all the resources at her disposal to (1) build a digital learning community, especially given the lack of inperson interaction, and (2) navigate the intricacies and nuances of online instruction, both of which are detailed in the following sections.

Learning community
Prior to the advent of COVID-19, the course was already offered in a blended format. Students used Blackboard to access each module and attended class in person, having f2f interaction with the instructor and peers. Students could also reach the instructor and peers using QQ and Blackboard outside class hours. It was a blended learning community. This sense of community became especially critical in the virtual learning context. The lack of physical presence necessitated more instructional initiatives designed to connect students. To build and nurture an online learning community in this course, three distinct online platforms were utilized.
The first platform was Blackboard, through which students had access to course materials and other learning resources prior to class. Since Blackboard was required by the university, 16 modules on the course webpage used for the former blended course format could then be entirely converted to online learning. This online course used the discussion forums on Blackboard to promote mutual learning, which was quite integral in building a dynamic learning community.
The second was QQ, a communication platform for students in and out of the classroom. The students needed to know where they could garner assistance from both peers and the instructor. QQ, China's popular instant messenger was selected to offer students synchronous communication with prompt feedback and asynchronous messages. Additionally, Tencent Document, an online word processing feature embedded in QQ, was used for peer editing.
The third platform was Zoom, which was crucial in building a virtual learning community. Zoom allowed synchronous learning, where the instructor could live stream to teach virtually. This course required live student discussions and presentations, necessitating synchronous class meetings. Zoom was chosen over Tencent Meeting for virtual sessions due to its breakout room feature. This function facilitates small group discussions, where the instructor could split the class into small groups and enter each room to oversee student discussion. Furthermore, Zoom was also used for students to conduct individual meetings with the instructor after the final exam, as each student needed to schedule a conference to go over their final video.
Moreover, the instructor strove to help students keep a propitious attitude towards online learning by creating a welcoming atmosphere and cultivating a positive outlook towards online learning. The instructor made a welcome video to introduce herself and the course, while communicating with students her concerns and expectations for student performance. The instructor noted the challenge of moving the course completely online and hoped that "public speaking" would not be reduced to "private talks." Students were encouraged to take full advantage of the online learning experience. Though students were not able to attend in person, practicing virtual presentation skills would prepare them for future job interviews and job-related requirements in the digital age. Both the instructor and the students learned to be flexible in their response to the unexpected interruption of traditional teaching and learning.

Instructional complexity
The transition to online teaching required the instructor to adapt the course content, delivery of that content, and assessment of learning. It compelled the instructor to have a holistic view of teaching and take a proactive approach in considering learner needs in course design from the outset. Clarity in instruction was essential, which is described in the following subsections of online modules, class management, assignments, and assessment.

Online modules
The course, which ran for 15 weeks, consisted of 14 modules addressing weekly course content with two additional modules for the midterm and final exam preparation. The content of the modules included topics such as using language effectively and speaking confidently, rhetoric and ethics of public speaking, informative speech, persuasive speech, impromptu speech, and critiquing speeches.
The instructor tried to present the online course content in a clear and accessible way. For example, Week 1's module provided an inclusive design of introducing the overall framework of this course. The instructor utilized visuals to inform students about the course outline in Week 1's course orientation. Students could learn how three stages of learning were involved in the virtual flipped classroom model (see Figure 1). To avoid overwhelming students with three distinct online platforms, the instructor presented students with a summary indicating the purpose of and how information was presented in each platform (see Figure 2). Furthermore, the instructor made a video to go through how the students could navigate Module 1.  Furthermore, each weekly module was adapted to online learning by reinforcing the value of utilizing the preview section. Before the COVID-19 pandemic, online modules were mainly used for students to preview the content that would be discussed in class meetings. In the preview section, the students viewed video lectures and read about that lesson's main topic (see Figure 3). The instructor could check how well students learned in the preview section by asking a few questions and observing students' discussions during the lesson in the classroom. However, the transition to online teaching made synchronous observation of students' responses to the preview content exceptionally difficult, which might have caused some students to be less motivated in previewing and hampered the efficacy of flipping the virtual classroom. Therefore, the instructor included each module's preview in the assessment to encourage students to probe further into the weekly course content. Students were required to finish the weekly preview and write a discussion post to answer the assigned questions. This feature involved giving credit for discussion posts after reading the previews for each module, except those for Week 1's module of the course introduction, Week 2's module, midterm, and final. Week 2's discussion post was not graded to give sufficient time for students' acclimation to the new online course. The twelve posts starting from Week 3 were worth one point each for a total of twelve points. In addition to instruction of the weekly module and the weekly Q&A forum, each module was arranged chronologically with three main components: preview (before class), in-class discussion, and the assignment (after class) (see Figure 4).

Class management
The lack of f2f communication made class management in a virtual class extremely critical. First, in order to organize the class effectively with clear directions, the instructor developed a set of class routines. For instance, the instructor informed students of the weekly teaching sequence. In each module, learning objectives were clearly stated in the instruction section. Every week, an announcement about important learning tasks was posted along with an updated module. Additionally, the instructor posted a class meeting agenda (see Figure 5) in each module both on Blackboard and QQ before each class meeting. Next, in order to promote student engagement in this course, weekly group discussions and presentations became course routines. Before the pandemic, class time was mainly devoted to discussions and presentations. Following the move online, how the classroom could be flipped virtually largely depended on whether the online platform could sufficiently provide access to group discussions. In each class, the instructor assigned students to small groups for collaboration by effectively utilizing the breakout room feature. Each group then presented back to the whole class after discussions within their groups (see Figure 6).

Figure 6.
A group presentation featuring screen sharing on Zoom (The picture shows the first author and three students in a breakout room. Informed consent has been obtained from the three students in the picture)

Assignments
Giving assignments in online teaching is a daunting challenge. Without f2f communication, precise and unambiguous language was essential in providing clear instructions about assignments. Throughout this course, the instructor also employed multiple media to give students additional reminders about their assignments.
The students were notified and repeatedly reminded of their assignments via six different means. First, on the syllabus, there was a table showing the deadlines of all assignments. Second, the instructor designed a weekly schedule where weekly themes, learning activities, and assignments were listed. Third, when students previewed each module on Blackboard, they could access an assignment section with clear instructions about the weekly assignments. Fourth, in weekly class meetings, the instructor dedicated one slide to reminding students of each week's assignments. After class, a screenshot of this assignment slide was also shared on QQ as a fifth way of announcing homework. Apart from all the above-mentioned reminders, an updated announcement greeted students upon entering the course website and alerted them to the weekly assignments.

Assessment
The evaluative demands of a public speaking course make it more suited for the f2f format, meaning a f2f format allowed proximal, moment-to-moment guidance and evaluation of speech. Transition to online teaching required the instructor to align assessments with the new changes in this virtual flipped public speaking class. The changes are described below.
First, a lack of physical f2f communication made it difficult to assess students' participation in class and small-group discussions. As a result, student participation was assessed in terms of their engagement with previews. The instructor added one preview section to each module, requiring students to write a post about assigned readings, recommended talks, and the video lectures.
Second, the students were encouraged to prerecord their midterm and final speeches to avoid any disruptions caused by unexpected technical glitches. They were required to show their faces as well as their visuals in their videos. The students' prerecorded videos were blended with a live Q&A. They were also allowed to give live presentations and use their recorded speech videos as backup.
Third, the rubric used for physical performance in speech was revised to allow the instructor to assess virtual presentations. Nonverbal communication, such as hand gestures and physical movement, was not included in the rubric. The criteria for grading body language were revised to evaluate eye contact and facial expressions instead.

Methodology
In this section, we discuss the epistemology and the theoretical perspective embedded in the study and then present the research questions. Next, we describe data collection and data analysis for this study.

Research questions
Since the study attempted to explore EFL undergraduate students' perceptions of their learning experience in a virtual flipped class integrated with MOOCs, it is important to understand how they developed their knowledge and understanding through experiences and interaction with others in the learning process. Thus, the epistemology for this case study was constructionism, where "[m]eaning is not discovered, but constructed" (Crotty, 2004, p. 9). Indeed, constructionism suggests meaning is synthesized from human engagement with the world. Along those same lines, interpretivism, which advocates for the unique construction of meaning for each subjective individual, was the theoretical perspective to further conceptualize students' interpretation of their experiences in this public speaking course.
In this case study about the students' virtual flipped learning experience with MOOCs in an EFL context, the following two research questions were asked: 1. How did the EFL students perceive their own experiences with MOOCs and virtual flipped learning in the English public speaking course? 2. How did the incorporation of MOOCs and virtual flipped classroom model facilitate student learning in this online course?

Data collection
The study corroborated results from multiple data sources, including surveys, focus group discussions, videos and feedback on students' speeches, peer evaluations, and the instructor's reflective teaching journals written over four months. Of the 30 students who enrolled in this course, there were three international students, one from Latvia, another from Lithuania, and the third from Cambodia. Due to the linguistic diversity, the survey was written bilingually in both Chinese and English with 49 questions to investigate students' learning experiences in this online course and their feelings towards the course design, including the flipped class model, MOOCs, and the three online platforms (Zoom, Blackboard, and QQ instant messenger). Several open-ended questions gave students free rein to elaborate on their opinions, which could also be in Chinese or English. For example, if students' response to Question 25 about whether the three online platforms fulfilled their learning needs was negative, they would find a follow-up question prompting them to explain why. The instructor distributed the survey online after the last class. Students completed anonymously so they could express their views freely. Twenty-five out of 30 students completed the survey. In the received surveys, two were completed in English and the other 23 were in Chinese.
The initial analysis of the survey results was used to develop focus group discussion questions. A focus group invitation was sent out after the course had officially ended and all the grades had been finalized. Five students, three males and two females, volunteered to participate in the discussion. The interview was conducted in Mandarin and lasted approximately 90 minutes. The interview was recorded using Zoom, transcribed, and translated into English. There were thus two sets of data.

Data Analysis
Charmaz's grounded theory (2006) was employed in analyzing the responses to the open-ended questions in the focus group discussion to glean a constructivism framework for conceptualizing students' understanding. A grounded theory is a theory "derived from data, systematically gathered and analyzed through the research process" (Strauss & Corbin, 1998, p.12). Grounded theory "consists of systematic, yet flexible guidelines for collecting and analyzing qualitative data to construct theories 'grounded' in the data themselves" (Charmaz, 2006, p.2). This theoretical orientation focuses on data from participants to construct theoretical meaning, rather than fitting the data to some preconceived theoretical framework. The purpose of applying grounded theory to this study was to develop a theory about EFL undergraduate students' perceptions of their learning experience derived from the data itself. Analysis of the data via a rigorous coding process demonstrated students' active construction of their own experiences with MOOCs and flipped learning.
The survey outcomes corroborated results supplied by the focus group discussion participants, which informed the initial coding categories, confirming the parameters of our inquiry. All data were analyzed through four stages: initial coding, focused coding, axial coding, and theoretical coding. In the initial coding stage, we started coding the two sets of data by finding recurrent themes operationalized by their frequent occurrence in our data: the survey questions and the focus group interview. The process involved word-by-word, line-by-line, and sentence and paragraph analysis to identify recurring keywords in the transcripts. Then we engaged in focused coding by compiling keywords from the two sources (survey and focus group interview) and comparing our data to pre-established codes from the literature review, like the types of instructional format and associated affects, such as enthusiasm, apprehension, frustration, and neutral. In the third phase, axial coding, we worked on categorizing and subcategorizing the codes and looked for tentative links between the categories and subcategories. For example, by categorizing the codes tied to instructional format using MOOCs, flipped classroom model, and three online platforms (Zoom, Blackboard, and QQ instant messenger) both in the survey and the focused group discussion, we found students tended to reflect on how they strategically adapted to the new experience of engaging with this course online. Thus we used "adaptation" to indicate students' exploration of the instructional format by addressing both the opportunities they enjoyed and the challenges they faced in the learning process. Finally, three categories of codes (adaptation, engagement, and accountability) were generated from the focused codes in previous phase. Then, we created several worksheets in EXCEL to refine subcategories into the main three categories (adaptation, engagement, and accountability). Table 1 below shows main categories and how subcategories were linked to each category in axial coding. Additional data, such as student speech videos and feedback, peer evaluation forms, and the instructor's reflective teaching journals, were all used to corroborate and confirm preset codes identified via the initial coding process using grounded theory from surveys and focus group interview. Thus, the coding process was iterative in that data was checked for additional codes multiple times for ones we may have missed prior to axial and theoretical coding (Finlay, 2013). Collaboration in the classroom (discussions, presentations, and peer evaluations) Collaboration out of the classroom (discussions forums, and assignments) Peer interaction (breakout room discussion, peer evaluations, and discussion forums) Individualized attention (instructor feedback and student teacher conferences) Accountability Preparation (watching video lectures, readings, note taking) Participation (discussions, presentations, peer evaluations) Improvement (awareness, confidence, course content-e.g. verbal & nonverbal language, visual aids) Figure 7 illustrates the four coding phases. The dotted line means that the mutual relationship between survey questions and focus group data in the initial coding is less emphasized as in the focused coding, where the solid bidirectional lines indicate coding of responses to survey questions frequently inform coding of focus group responses, and vice versa.

Figure 7. Flow of coding process
In first three coding phases, we revisited and repeatedly compared the data with each other, codes with other codes, and codes with the data for saturation of data categories. Axial coding led us to further conceptualize "how the substantive codes may relate to each other as hypotheses to be integrated into a theory" (Glaser, 1978, p. 72). In the fourth and last phase, theoretical coding, fractured concepts generated in the previous coding process were weaved into a theoretical scheme. On the basis of Table 1 created in axial coding and all the substantive codes generated in previous stages, a theoretical scheme (Chiodi, 1991), or the connections between different concepts central to this phenomenon, was created by tying codes with the research objectives. This theoretical scheme is depicted in Figure 8 in the next section. Zhuo Li and Xiaodi Zhou

Discussion of Findings
The discussion below is organized according to the theoretical scheme derived from the coding. Results of data analysis indicated that students were generally positive about their virtual flipped learning experiences integrated with MOOCs. Consistent with findings in other studies on student responses to flipped instruction (e.g., Chua & Lateef, 2014), all participants in this study preferred the flipped classroom rather than traditional teacher-centered instruction or learning strictly with MOOCs, which were the pre-pandemic instructional formats they experienced in other courses. The flipped classroom model that is more a key element in the current course allowed mastery of learning through practice in the virtual classroom and made the students feel increasingly connected to the instructor and their peers. Incorporating MOOCs in their flipped learning expanded students' learning opportunities to use all available intellectual and academic resources. In addition, collaboration was cultivated by cooperative learning in and out of the classroom.
Through an iterative analytical process informed by our coding, we found that students adjusted well to the wholly online instructional formats. They demonstrated, especially at the initial stage, adaptation to flipped learning and using MOOCs during COVID-19. The students were engaged in exploring a variety of digital platforms, learning resources, remotely collaborating with peers, and interacting with peers and the instructor. Accountability cultivated in this process further reinforced their engagement and enhanced their adaption to the virtual learning environment. In Figure 8, the arrows outside the oval represent reciprocal interactions among adaptation to the instruction delivery, engagement with the learning material, and accountability of learning course content.

Adaptation
In terms of adaption, students described both how they acclimated to using MOOCs, flipped learning, and the three specific online platforms when the classes were moved fully online.

MOOCs
Since the pandemic, online courses around the world have relied heavily on the use of MOOCs for instruction delivery (Wang et al., 2021). Thus, this medium will be discussed as the predominant mode of learning during COVID-19. Eleven of 25 students used MOOCs before taking this course, but only 4 students indicated they used MOOCs due to course requirements. The other 7 students had watched MOOC videos to supplement learning out of their own interest. The students reflected that the weekly online class meetings on Zoom necessitated following instructions to finish the MOOC video lectures on Blackboard. For example, one student reflected on his own MOOC learning experience: Without any external guidance or supervision, it would have been extremely difficult to learn independently using MOOCs.
The students welcomed the flexibility of using MOOC videos on Blackboard. They appreciated how they could watch and review video lectures at their own discretion. For instance, one student commented, "I don't have to worry that I can't understand the content and miss any important points. If I find one part of the video lectures is difficult, I can simply go back to watch that again." However, some students commented how it was easy to forget the content of video lectures. In the focus group discussion, students remarked that many failed to take sufficient notes of lectures.
The main drawback of MOOCs, as the students in the focus group pointed out, was the lack of direct interaction with each other in their engagement with the class. Hence, they still embraced class meetings where they could talk to and listen to peers and the instructor in real-time, responding to their contributions in dialogic synthesis. These findings correlate with other studies showing how MOOCs can complement and augment, but not replace, traditional education (e.g., Hlinak, 2016;Li, 2015).

Flipped learning
The survey showed only two of all the 25 students had prior experience with flipped learning. In the focus group discussion, all five participants stated that they had heard about the flipped classroom model before but that this pandemic-adjusted course was their first time actually experiencing flipped learning. They appreciated that flipped learning freed up time to have group discussions and presentations. It became especially important due to the social distancing necessitated by the pandemic that they connected with their peers through class engagements. They cherished the opportunities to practice speaking with each other. One student commented that the experience in this course was more like "actively learning" versus his previous experience of just "listening to lectures" from the instructor before the pandemic. He added that this course completely changed his prior conceptions of learning.
Meanwhile, the course format was challenging for those students who had been accustomed to passively listening to instructor lectures. For example, some students stated that flipped learning was more time-consuming when compared to the traditional lectures even though the learning outcome was much better. They added that they were "forced" to be more active in learning rather than passively listening to lectures. In such student-centered classes, students had to shift their previous role as docile recipients of knowledge to active participants in constructive learning. Previewing lessons became crucial, as those already primed before class expressed substantially more satisfaction about their learning than those who failed to do so. This is consistent with Çakıroğlu and Öztürk's finding (2017) that self-regulation is necessary for meaningful learning in flipped classrooms.

Three online platforms
The survey findings showed that all students agreed that Zoom was an effective platform for this course (see Table 2). In the focus group discussion, students compared their learning experience in this course with other courses where Zoom was also used. They recalled Zoom was used mainly for lectures and how the students rarely uttered a word in those other courses. In this course, however, the students felt they needed to stay focused the entire time because they were consistently expected to interact with peers and the instructor. Though the survey results showed not all students thought that Blackboard was effective (see Table 2), the participants in the focus group did not express any qualms about using Blackboard. They believed a platform like Blackboard was needed to share learning materials systematically and to structure learning. However, the platform's malfunction in Week 5 did cause students to lose confidence in its efficacy. Furthermore, 56% of the students regarded QQ, the main instant messaging platform in China, as an effective interface in communicating with peers and the instructor. In fact, when Blackboard was down, the instructor used QQ as a temporary solution to successfully deliver learning materials.
All students except for one believed that the three platforms could sufficiently fulfill their learning needs (see Table 3). 56% of students assessed the effectiveness of utilizing three online platforms as "very good" and 44% of them thought it was "good" (see Table 4). Students' positive response to the three platforms' application in learning is in line with Ismail and Abdulla's (2019) finding that the adoption of virtual flipped classroom can bolster teaching and learning efficacy.  In the survey, two questions specifically probed students' attitudes towards online learning (see Table 5 and Table 6), and findings suggest that the majority of students still preferred f2f learning when given a choice.  Not sure 16% (4) Focus group participants who preferred online learning expressed that previewing material online could sufficiently prepare them for class. Furthermore, the flexibility in online learning gave them more autonomy and investment in their education. One participant enjoyed the simulation of f2f interactions in the breakout room feature on Zoom, stating that "it is more convenient to have group discussions online, especially in displaying our own notes by sharing the screen." Another student expressed her positive feelings about the breakout rooms, saying that "discussing in the breakout room makes me feel like I'm in a meeting room so that we won't be bothered by other groups and I like the interaction whenever the teacher may have with us in our own 'room'." Those who insisted offline learning was more advantageous stressed the importance of maintaining human connection in learning through physical f2f interactions, even noting the importance of body language in public speaking. Nonetheless, they concluded that using all three platforms in tandem had indeed fulfilled their learning needs when inperson learning was not an option.

Engagement
The students reported that the traditional lecture style of teaching was the norm in their STEM courses before the pandemic, so this course represented an entirely new mode of learning. In their pre-pandemic courses, it was impossible to address all individual needs, so the students relished the individual attention from the instructor in the current public speaking class. With more class time devoted to group discussions, the instructor was able to listen in on each group, engaging meaningfully with every student.
We now analyze the course as an elective, and then with respect to digital engagement. This course was an elective so students voluntarily enrolled in this class. However, their low expectations of the course did affect their level of engagement. Chinese students commonly held lower expectations in elective courses than in their core courses (Li et al., 2005). However, as some students mentioned in the focus group discussions, they felt they had learned much more solid knowledge than they initially expected. Part of the reason was their preconceptions about elective courses being easier or less intensive. Their expectations of the course notably affected the approach they took in the learning process. One student's comment typified student attitudes toward elective courses in general: We usually select electives for fun or higher grades. Overall, per their end-of-the-semester evaluation of the course, students enjoyed using MOOCs as part of their learning, as they felt more connected to the larger academic community. They reflected on the sense of comradery built through collaboration in and out of the virtual setting, peer interaction that occurred in group discussions and peer evaluations, and the individualized attention from the instructor. Moreover, students reported that meaningful use of the three online platforms bolstered their engagement and investment in learning. They could use Zoom for live-video communication with peers and the instructor during class meetings. They also believed that utilizing one-on-one meetings with the instructor to discuss their final speeches was beneficial in facilitating their speech improvement. One troubling occurrence in these digital meetings was how most students did not show their faces either in breakout room discussions or in whole class meetings. The showing or masking of their appearances related to perceptions of their online learning in using Zoom. In f2f classes, students are not given this option of hiding their appearance, and so the ability to veil themselves from others during online class became a factor impacting their experiences utilizing this interface. By surveying students about their attitudes regarding showing their faces on Zoom, we target their perceptions of their digital learning experiences (see Table 7). We found almost 50% of the participants expressed their willingness to show their faces on Zoom, yet very few actually showed themselves in class when they were not required by the instructor. In the focus group discussion, all five students agreed that seeing each other online could help foster personal connections and a sense of community. Nevertheless, often times, they did not feel comfortable showing their faces because they worried about their appearance or that their backgrounds were unsightly. One student did mention, however, that it was easy to enable virtual backgrounds instead if he did not like the real backgrounds. Overall, they appreciated that the instructor encouraged but did not mandate they should show faces on Zoom. Student engagement was also tied to their interaction and collaboration on Blackboard and QQ. On Blackboard, they were required to give constructive feedback, peer evaluations and course reflections. Peer collaboration connected students as they worked in and out of the classroom. In class meetings, they discussed with each other in breakout rooms and presented to the class in groups. They related to each other because they needed to make posts as a group and give each other peer feedback. Aside from the assignments, they also worked together on group presentations for their midterm. QQ was the most utilized communication tool throughout their learning process. All of their comments reflected that the learning community constructed through the three online platforms in tandem kept them engaged in attaining common learning goals via the different learning structures.

Accountability
In this virtual flipped learning context, "accountability" emerged when students took the initiative in learning and demonstrating responsibility for that learning. Students began exhibiting accountability once they adapted to the virtual learning environment employing the flipped classroom model and MOOCs. According to survey results and the course evaluations, they were engaged with course materials, collaboration, and interaction with peers and the instructor. Meanwhile, their responsive learning reinforced their engagement and enhanced comfort in the new learning environment.
Via careful course planning, student accountability for learning was fostered through intentionally scaffolding the course expectations. Considering that the students might take some time to adapt to the module format on Blackboard, the instructor began grading preview postings in Week 3. With two weeks of "acclimation time," students learned that previewing was essential due to in-class discussion prompts that necessitated knowledge gained from previewing. Expectations for before, during, and after each class meeting were clearly reinforced. Once they became more confident with communicating in class, they noted their own improvement and became more engaged in interacting with peers and the instructor. For example, one participant explained how his accountability was fostered in learning this course: I realize that I had a problem in preparing presentations before. I always recited the manuscript but tended to forget when I got on the stage, which made me very nervous. In taking this course, we always discussed in groups first and had to share our discussion to the whole class on Zoom. I find I have formed a habit to come up with a quick structure before presenting to the whole class and started to feel free to add some specific messages in delivery. Gradually, I became more active in previewing every week's content on Blackboard because I just wanted to get more ready for group discussions and presentations.
This student explained how his adaptation to the course format (i.e. "always discussed", "had to share," "formed a habit") made him more engaged (i.e. "feel free", "more active in previewing") as he became more confident in speaking, which also reinforced his accountability (i.e. "wanted to get more ready") to complete assigned preview questions. He also showed meta-awareness of his own learning process and his acclimation to this mode of instruction.

Conclusions and Implications
While the virtual flipped classroom model integrated with MOOCs posed pedagogical challenges, it also offered opportunities for exploring learning alternatives, embracing innovations, and fostering flexibility and resilience in response to emerging instructional needs. This study suggests that students' adaptation, engagement, and accountability all reinforced the process of tackling the challenges in the new learning format. Students' adaptation in learning derived from their level of engagement fostered in the dynamic learning community. According to student feedback, their engagement benefited from their adaptation to the course, as well as accountability for their learning. Accountability that occurred concomitantly with adaptation and engagement could be further developed in the learning process. The current educational disruption has made us the authors rethink the significance of human connection in traditional classroom-based education and the role of online teaching following the pandemic. Admittedly, tackling the challenges in online teaching and learning made both the instructor and students realize how essential physical presence and f2f interactions were in the organic process of teaching and learning, which should be a dynamic social engagement privileging human contact.
At the same time, such heavy reliance on virtual teaching required adapting technology to a new realm. Students who cannot physically attend class meetings or student-teacher conferencing may be able to participate digitally or asynchronously. Moreover, this study indicated that students were more prepared with the activation of previewing functions as utilized in this online course.
This case study may have some implications for ELT practitioners who need to convert courses that are dependent on f2f interactions to solely online delivery. First and foremost, careful planning and clear instructions are vital to avoid any ambiguity that may impede online learning. Meanwhile, given the plethora of MOOCs which can be used for English learning, EFL teachers can take full advantage of the massive repository of educational resources. Additionally, the flipped classroom model maximizes EFL learners' learning opportunities and practice time. Strategic use of online platforms in conjunction with MOOCs within the flipped classroom model can minimize educational disruption without diminishing academic rigor and quality, and even reframe a global disruption into a constructive pedagogical approach. Moreover, it is essential to supplement a primary platform with several others to compensate for each's shortcomings. For example, in our case, three platforms (Zoom, Blackboard, and QQ instant messenger) were used to satisfy the various needs of online class meetings, course materials delivery, asynchronous and synchronous communication. Next, collaboration with the technology department to effectively orchestrate a suitable instructional condition is needed for effective instruction and interim solutions should be offered when technical problems occur. As evident with the present case, there were unforeseen challenges due to technical glitches and student inhibitions, which perhaps necessitates greater coordination with the administration to create a learning environment more amenable to digital learning. Last, admittedly, the lack of f2f communication, though approximated by Zoom engagements, still did not build the requisite sociocultural context for learning (Vygotsky, 1978), especially one as dependent on human connections as public speaking. In order to have students stay connected with peers and the instructor, we may reach out to students by creating a friendly virtual learning environment to maximize interpersonal interactions.
Future studies could explore exactly how a well-designed hybrid model of learning can be conducive to more effective learning, connecting the different modalities to the specific results. Furthermore, individual student performances or instructor qualities could be more precisely assessed as a factor in effective digital instruction. Perhaps a larger sample size could provide greater external validity. Other linguistic and cultural contexts could also be addressed.