A grounded theory exploration of language massive open online courses (LMOOCs): Understanding students’ viewpoints

Massive open online courses (MOOCs) have been called the biggest innovation to happen in education for 200 years for their unique attribute of being open and free to any individual with Internet access; however, their high dropout rate has led many people to be concerned or dubious about MOOCs’ effectiveness and applicability. The applicability of MOOCs in English for speci�c purposes (ESP) courses (in this case, Hospitality English) needs more empirical evidence; the present study intends to help �ll this gap and extend our current understanding of this issue. This study followed a grounded theory methodology to develop a theoretical model based on constant dialogue between collected data and the literature. The results suggest that most participants had positive perceptions of LMOOCs in general but that some of them doubted the applicability of LMOOCs. Most participants said they would continue to use LMOOCs in learning, depending on the attributes of speci�c courses. Based on the extracted data, a conceptual model for the applicability of LMOOCs is proposed.


Introduction
Web-based learning has indeed engendered a brand-new paradigm for teaching and learning (Kop, 2011;Soares, Lopes & Vieira, 2015;Tiryakioglu & Erzurum, 2011).Scholars such as Jones, Blackey, Fitzgibbon, and Chew (2010) and Shadiev, Hwang and Huang (2015) postulate that students' analytic and interpersonal skills can be enhanced through socialization or cooperative learning in virtual environments (Joo & Park, 2015).Massive online open courses (MOOCs) are one innovative product of these technological and educational developments, in which course design mentality has shifted from being 'curriculum oriented' to 'experience oriented' (Freitas et al., 2015).Despite the fact that MOOCs have been used worldwide, the number of language learners who adopt MOOCs for learning remains limited (Uchidiuno et al., 2017) because only handful MOOCs are speci cally designed for language learning and teaching (Friðriksdóttir, 2018).In the eld of language education research, pertinent study on MOOCs is rather sparse (Sokolik, 2014); to extend our current understanding of the effectiveness of language MOOCs (LMOOCs), more empirical evidence is still needed to serve as a rationale for future application.
MOOCs have been described as 'the biggest innovation to happen in education for 200 years' (Cadwalladr, 2012) for their unique attribute of being open and free to any individual with Internet access, which allows the courses to be 'massive' in terms of the number of students involved (Alraimi, Zo, & Ciganek, 2015;Voss, 2013).However, some scholars, such as Romeo (2012) and Jackson (2013), have raised concerns about the actual applicability of MOOCs in education.High dropout rate is a major challenge for all MOOC designers (Daniel, 2012;Hood, Littlejohn, & Milligan, 2015;Jordan, 2014;Kolowich, 2013).Although it is unsound and unfair to evaluate the quality of a course by completion rate alone, high dropout rate still causes people to have second thoughts about MOOCs (Marcus, 2013;Reich, 2015;Strack, 2017).
In most MOOCs, instruction is usually delivered via video clips and some activities for peer interaction (Martín-Monje, Castrillo, & Mañana-Rodríguez, 2018;Soares, Lopes & Vieira, 2015).In this sense, it can seem as if MOOCs have simply digitalized instruction and delivered it through the online platform without devoting any consideration to how to encourage learners to participate continuously or how to enhance their motivation for learning (Kizilcec, 2013).Even so, MOOCs' affordances-breaking down limits of time and space and bringing in learners with various cultural, language, and geographic backgrounds within one class through the Internet-can possibly realize social learning and enable learners to more effectively construct their own knowledge.For this reason, it is a topic worthy of researchers' attention and turning MOOCs into effective tools will enable every learner who is eager to learn has access to learning materials (Hood et al., 2015;Veletsianos, Collier, & Schneider, 2015), including LMOOCs.This present study was designed to explore the applicability of LMOOCs for EFL learners to learn Hospitality English.Three research questions are proposed to be addressed: 1. How do learners in Hospitality English course perceive their learning experience with LMOOCs? 2. What are the underlying reasons of their using LMOOCs in their Hospitality English learning?
3. What are the biggest bene ts and challenges they have encountered in their experience using LMOOCs in hospitality English learning?
To answer these questions, this study rst reviews pertinent academic works.Afterwards, the research methodology is reported and the results of this present study are presented.Then, based on the results, a conceptual model is developed, and suggestions for future research and teaching practice are proposed.A brief conclusion follows.

Literature Review
Theory of Social Learning and MOOCs It has been postulated that human beings can learn new things through observing, understanding, and imitating (Saputro et al., 2019).
The theory of social learning was proposed by Albert Bandura (1977); it posits that learning is a process wherein learners trigger learning mechanisms through observing and imitating others' behaviors or interactions, acquiring new knowledge accordingly.In other words, social learning points out the importance of interactions between peers, within the environment, or both for the initiation of cognitive behavior (Tu, 2000).Based on this theory, social learning aims to cultivate an individual's active learning habitus, so that new knowledge will no longer be a one-way infusion but will be constructed through interactions (Brown & Adler, 2008).The concept of social learning can be suitably implemented via MOOCs in situations when learners are able to learn through observing individual or social interactions that respond to the model built by the instructor (Spencer, 2015).
With respect to learners' engagements in MOOCs, Hill (2013) identi ed ve different behaviors that learners of MOOCs perform.According to their level of engagement, they are no-shows (who register but never log in), observers (who log in and read the materials but never take part in any activities), drop-ins (who occasionally participate in activities), passive participants (who routinely log into the course but do not partake in activities) and active participants (who fully engage in the course).Furthermore, Anderson, Huttenlocher, Kleinberg, and Leskovec (2014) analysed learners' engagements patterns in MOOCs with a set of large-scale trace data and proposed a taxonomy of ve behavioral patterns corresponding to ve types of learners: viewers (who view the MOOC materials and submit few assignments), solvers (who submit assignments for grades but seldom view the lectures), all-rounders (who keep balance between viewing the materials and submitting assignments), collectors (who download materials from MOOCs but submit few assignments), and bystanders (who register but do not often engage in course activities).The researchers also emphasized that an individual's engagement in a MOOC is different from engagement in the o ine world and stressed that further exploration by researchers on the underlying reasons for this disparity were necessary.
In a MOOC, learners with various backgrounds can be placed on the same platform; empirical evidence indicates that if a class is composed of students with different personality traits and learning styles, this will promote learning motivation (Miller & Lu, 2003;Clerehan, Turnbull, Moore, Brown, & Tuovinen, 2003).Furthermore, such socialization enables learners to fortify their knowledge acquisition (Thomas, 2003), as captured in Vygotsky's 'scaffolding' concept, which emphasizes the importance of interactions (Peterson, 2009).This implies that pedagogy should be focused on the issues of why and how an individual learns instead of what he/she is learning (Ruey, 2010).Importantly, even if learners are not involved actively in the learning process, meaningful learning may still take place (Goodyear, 2005).Speci cally in the context of MOOCs, learners can possibly match their passive learning styles and how they internalize learning content through passively meaningful learning (Murugaiah, Atan, Samsudin, & Idrus, 2004).

MOOCs for Language Learning (LMOOCs)
The term language MOOC, or LMOOC, was coined by Bárcena and Martín-Monje (2014) to indicate 'dedicated Web-based online courses for second languages with unrestricted access and potentially unlimited participation' (p.1).As Ding and Shen (2019) pointed out, LMOOCs are different from other MOOCs in two aspects.First, LMOOCs offer a greater variety of content such as topics and learning activities as well as materials for learning.Another difference concerns the role played by instructional videos in LMOOCs; rather than merely offering pre-recorded lectures, videos serve as one of the vital language sources with authenticity for learners' socialization.Perifanou (2015) also argued that social interactions and collaborations among learners can be promoted with LMOOCs.Because of these special attributes of LMOOCs, learners using them have a varying degree of engagement in them (Ding & Shen, 2019).For example, LMOOC learners tend to disengage selectively from some topics or activities even though they are highly autonomous.
According to Beirne, Mhichíl, and Cleircín (2017), there are 143 LMOOCs offered worldwide and the number is still growing.LMOOCs have been deemed as 'the most attractive, criticized and polemic of all types of online language courses' (Martín-Monje, Read, & Barcena, 2017, p. 11), and they have attracted increasing attention in the eld of language education (Fuchs, 2017;Perifanou & Economides, 2014).Previous academic works on the implementation of LMOOCs report both positive and negative comments by students.The most frequently mentioned bene ts of LMOOCs include their exibility and the stress-free environment that they create.Also, learners' communicative competence may be enriched through interactions with others and LMOOCs are posited to be suitable to be the platform (Panagiotidis, 2019).Nonetheless, it still seems to be challenging to create adequate socialization in the virtual environment (Schilze & Scholz, 2018).Moreover, as Bárcena et al. (2014) have pointed out, language learning is a rather complex because 'it is both skill-based and knowledge-based' and various genres of capabilities would need to be delicately included in LMOOCs; hence, foreign language may not be an ideal subject for MOOCs.There are still many obstacles that need to be overcome by practitioners (including instructors, administrators, and students) of LMOOCs (Martín-Monje, Bárcena, & Read, 2013) and more empirical evidence are needed to provide suggestions for the practices of using LMOOCs in language learning and teaching (Martín-Monje, Castrillo, & Mañana-Rodríguez, 2018).

Research context
The LMOOC of this study was designed for anyone who was interested in learning English for a speci c purpose, i.e., learning English to work in the airline industry.The title of the LMOOC was in fact 'English for Airline Cabin Service'.The information about this LMOOC was disseminated through the networks of National Kaohsiung University of Hospitality and Tourism (NKUHT) in Taiwan; additionally, some private aviation training schools also helped to circulate the information to their students.This course was designed and implemented by NKUHT, where the delivery mainframe was housed.This LMOOC of English for Airline Cabin Service consisted of three major components: the video, grammar, and test (please refer to Fig. 1 to 4 for details).
In addition to these three parts, a forum was also created for interactions and socialization as suggested by Martín-Monje and Barcena, (2015).Learners might raise questions about the course contents and their peers who knew the answer could make responses.The design and pilot test ran for four months and had good results (e.g., the contents of this LMOOCs were in line with the goal of this course, smoothness of the platform); the course was then ready for learner registration.The course went into operation in February 2017; after three months of operation, focus groups were convened to collect qualitative data to address the research questions.The course lasted for three months, and the last round of interview was conducted in September 2017.

Grounded Theory
Given both the innovative platform for language learning that LMOOCs provide and the need to address the gap in the literature described above, grounded theory is the most appropriate methodological choice (Karpouza & Emvalotis, 2019;Noble & Mitchell, 2016).Grounded theory is de ned as 'the discovery of theory from data systematically obtained from social research' (Glaser & Strauss, 1967, p. 2), which makes it a preferable research method to re ect the ever-changing reality of psychosocial processes (Aldiabat & Navenec, 2018).Using the methodological approach known as grounded theory would enable the revelation of some insights about EFL learners' viewpoints on the applicability of LMOOCs in English language learning.Detailed information about the implementation of grounded theory is reported as follows and the process of data collecting, coding and concept generation is presented in Figure 5.

Participants
In keeping with grounded theory, theoretical sampling was administered to 're ne ideas, not to increase size of original sample' (Charmaz, 2000, p. 519).With this goal in mind, three focus groups of ve members were recruited (following the suggestion of Nyumba et al., 2017), for a total of 15 participants.There were 189 learners who registered to take this LMOOC but only 21 of them completed this course with a completion rate of 11%.These participants were chosen based upon four conditions: geographic location (northern, central and southern Taiwan), gender (three female and two male participants for each group), their total login time, and login frequency.It was the aim of the present study to understand the underlying reasons for EFL learners' willingness to use or not use LMOOCs; hence, one-third of participants' login time fell within one standard deviation of average length of time (6.3 hours) whereas the remaining two-thirds were equally above and below the interval.Focus group interviews were selected as a means of collecting data because of the group dynamic (i.e., focus group interviews) they give rise to, which may be able to generate more insightful, richer data in a shorter period of time due to social interaction among the group members (Krueger & Casey, 2014).These three groups were formed according to the three metropolitan areas (from northern, central, and southern regions of Taiwan respectively) where most participants resided.All the participants consented to take part in the interviews, and were informed that they could withdraw from the research if they felt uncomfortable carrying on and no penalty would be imposed on them.Detailed information about focus group interviewees is reported in Table 1 below.More importantly, 'the number of contacts with each participant and the length of each contact' (Onwuegbuzie & Leech, 2007, p.117) were also captured by the present study.After the participants were selected and their informed consent was collected, each of them was a code name based on demographic characteristics.For example, a female student from northern Taiwan assigned number two was coded as NF2 and a male student from southern Taiwan with number 3 was SM3.Additionally, since data collection and analyses were reciprocal, multiple rounds of interviews took place, and therefore another numerical code was given to each piece of collected data according to which of multiple interview sessions it was collected in.Thus, a piece of data collected from a male student, assigned number 1, from central Taiwan in the rst interview would be CM1-1.
Interview questions Chenitz and Swanson (1986) suggested that unstructured interviewing is the norm of data collection in the grounded theory which was adopted by this present study to acquire data which were not "uncontaminated" by existing theory (Cunningham & Carmichael, 2017).Interview questions of this study were mainly derived from three proposed research questions but detailed questions were generated based upon interviewee's sensitivity towards interviewees' respondents.According to the principle of theoretical sensitivity, which plays a signi cant role in grounded theory (Orland-Barak, 2002), researcher(s) should enter the research site with profound knowledge about the issues but without preconceived notions (Weed, 2009).Interview questions in this study were developed based on a thorough review of relevant literature.Sample questions include 'What do you like and dislike about LMOOCs the most?' and 'What is your experience of using LMOOCs?'However, the iterative nature of grounded theory required researchers to move between data and emerging theoretical categories; therefore, interview questions would be revised according to the interviewees' responses or follow-up questions were raised to clarify some vague points as well as raised depending on the interviewees' responses (Lawrence & Tar, 2013).

Data collection
This research adopted Ground Theory Analysis (GTA) to elicit insightful information about students' viewpoints on the applicability of LMOOCs.To ensure the quality of a grounded theory study, which is conducted on the premise of rich, empirical, original and trustworthy data that can re ect complex psychosocial process (Aldiabat & Navenec, 2018), a principle of theoretical saturation was adopted in data collection, meaning collecting data until new data do not add more insights to the current core category (Aldiabat & Navenec, 2018;Charmaz, 2006;Glaser & Strauss, 1971;Morse, 2004;O'Reilly & Parker, 2012).
In this study, a series of focus-group interviews were administered until theoretical saturation was met.Two moderators presided over the group interviews to facilitate group interaction and pace the process so all the participants would feel at ease to share their viewpoints.With the consent of all participants, interviews were video-recorded, and interview transcripts were created based on the video-recordings.Since the transcripts of the rst interview were coded, memo-writing was used to depict the relationships among concepts.Memos in this context are sense-making tools that inform of us the underlying meaning of codes and further move them into the conceptual level by linking similar points (Lawrence & Tar, 2013).Detailed information about the iterative process of grounded theory adopted by the present study is given in Figure 6.In total, four rounds of interviews were administered to reach the data saturation.

Data coding, categorisation and theory building
Data coding process is the most important task to construct meaning from the collected data in the GTA (i.e., grounded theory analysis) (Williams & Moser, 2019).Three steps are administered to eventually build theory (Yan & Horwitz, 2008), namely, thematic analysis, generation of a nities or variables, and interrelationship digraph analysis.In the phase of thematic analysis, the process of identifying patterns within qualitative data is undertaken to discover the theme and further make sense of it (Maguire & Delahunt, 2017).Moreover, three levels of coding should be administered (Yan & Horwitz, 2008) and constant comparative analysis is adopted for coding and category development through inductive processes (Tie, Birks, & Francis, 2019).Level 1 coding is breaking the transcribed verbatim into small basic yet meaningful ideas which leads to Level 2 coding through comparing these small units of ideas and pattern would be identi ed accordingly.Since interview questions of this research were designed to address speci c research questions, theoretical thematic analysis rather than inductive thematic analysis was adopted (Braun & Clarke, 2006).Level 3 coding mainly focuses on systematically summarising and organising abstract constructs from themes extracted from Level 2. Throughout the whole coding process, Williams and Mosers' (2019) advice that the researchers should "read and re-read interview transcriptions, eld notes, and associated data sources involved in the data collection searching for thematic connectivity leading to thematic patterns."(p.49) was followed.
Once the abstract constructs are recognised, the stage of generation of a nities or variables is initiated and thematically de ned clusters can be afterwards labelled as a nities which are elicited through a series of delicate examination on logic and meaning of analysed themes.The last step of GTA is interrelationship digraph analysis and the interrelationship between a nities is linked through cross examinations on the associative connotations and the direction of their relationship (i.e., which one is the driver and which one is the outcome) is determined based on the participants' comments (Yan & Horwitz, 2008).However, these steps were iterative as Charmaz (2008) put it "Data collection, analysis and resultant theory generation has a reciprocal relationship...it requires a constant interplay between the researcher and the data" (p.47).The application of the grounded theory for theory building were undertaken in the following steps which was in line with the method adopted by Soares and Oliveira (2019): 1. Transcribing the interviews, organized EFL learners' feedback about LMOOCs, 2. Identi cation of the codes based on the interviewees' feedback, 3. Warranting de nition of the main network between the codes, 4. Identi cation of new codes for pros and cons of using LMOOCs for ESP learning.

Results
RQ 1: How do learners in Hospitality English course perceive their learning experience with LMOOCs?
Out of the 15 participants in this study, 12 considered their experience learning with the LMOOC to be positive and suggested that the LMOOC was a convenient, helpful learning platform for them.They considered that their learning experience with LMOOCs and other MOOCS had been similar.Most of the positive statements on both types of MOOC were collected in the rst or second focus group interviews.For instance: I like to use LMOOCs very much.[A LMOOC] is more convenient then other e-learning platforms because it has more functionalities (e.g., multimodal materials and opportunities to interact with other learners) than other e-learning platforms that I have used… (NM1-1) Another participant in the same focus group supported this statement from NM1 right after hearing it, and, further, raised some intriguing points which stemmed from his experience learning subject matter with the MOOC: [O]h yes, I totally agree with him (NM1), the convenience of the MOOC is the main reason for me to use it.Nevertheless, there are more attributes which attracted me to use it.These attributes included [that it is] multimodal, open, easy to access, and more importantly, it ts my unstable learning time, because I had a job at a hotel and I had to be on duty at different times… (NM2-1) Participants in other focus groups expressed similar views.Most made positive statements about MOOCs in their rst focus group meeting; however, from the second interview on, some participants (7 out of 15) started to express differences between LMOOCs and other MOOCs.For example, participants FC4 and FS3 argued that they preferred in-person interaction because the service industry is a 'people industry' and it is quite important for prospective workers in this industry to have interpersonal skills; equally important, interactions would also be critical for language learning, particularly for communicative competence which [participants argued] would be hard to enrich with LMOOCs.
I have to admit that I like the exibility of MOOCs.However, I found out that not being able to physically interact with others, was the greatest drawback which could not be compromised [for] with convenience… (CF4-2) Another participant expressed a similar opinion at another focus group meeting, but tackled the issue from a different angle.
[T]he more I used MOOCs I gradually realized something that MOOCs just could not function very well, which referred to the warmth and affection which positive teacher-student relationships may harness to learning... (SM2-2) Albeit SM2-2 did not explicitly or directly mention that LMOOCs cannot provide the same bene ts as physical interactions for language learning, he did point out that what has been missing in MOOCs (regardless whether LMOOCs or MOOCs of other subjects) was teacher-student relationships.Another participant pointed out the concern that human learning might be reshaped by the technology.
I enjoy the convenience and usefulness of MOOCs, but I just feel that I should not be conditioned by computer… (CM5-3) Here, another downside shared by LMOOCs and other MOOCs was raised by CM5-3; participant CF4 agreed with CM5's statement and followed up: …I have to say that I have been kind of conditioned by MOOCs.Explicitly, I opened another window for the MOOC when I was at work and checked if there was anything going on or new feeds there.I don't know if it is good or bad, but it just turned out to be like this… (CF4-3) In summary, participants shared their viewpoints about their special learning experience with the LMOOC that differentiates it from other MOOCs.The major difference was not being able to enrich language learning through face-to-face interactions, a drawback that cannot be compromised by the convenience of MOOCs.As for similarities, participants started with positive perceptions of both, mainly focused on convenience and exibility.Nevertheless, after a certain period of time, some participants had concerns about (L)MOOCs and worried about being conditioned by them, which meant that their lifestyles had somewhat been affected by (L)MOOCs.These participants tended to be too reliant on the technologies and this situation inferred that they would possibly be less effective learners in traditional classroom instruction, where the ecosystem would be different.
RQ2: What are the underlying reasons of their using LMOOCs in their Hospitality English learning?
After identifying participants' perceptions of using the LMOOC, this research further aimed to understand the underlying reasons why they did or did not intend continue using it.In various focus group interviews, participants who held positive perceptions focused on the convenience and on the diversity of participants the LMOOC was able to convene.
It is indeed very convenient to use LMOOCs to learn.On the other hand, some of my natural born abilities; e.g., my working memory capacity and attention span, seemed to vanish because of the convenience of MOOCs…I did not need to remember some details taught by the instructor because all I had to do was to re-play the video clip or went back to the previous page… (SM2-2) [O]ne of the impressive affordances of LMOOCs is that I had the chance to interact with others who I did not know and I felt really comfortable about it… (CF1-2) Another participant pointed out that she generally appreciated the affordances of MOOCs but did not think it was a good idea to learn English via LMOOC: I will continue to use MOOCs to learn different subjects but not hospitality English because MOOCs lack one of the most essential aspects of hospitality English-face-to-face interaction… (SF1-2) I believe that I will use LMOOCs in my future study of hospitality or English courses because it is really convenient to me and to me, exibility is everything because of my unusual schedule… (NM2-1) Another participant, who did not speak often in the interviews, also commented on this question.Although she had been shy in the interviews, she paid close attention to the others' discussions.She raised the following point: I like LMOOCs and I will keep on using them in hospitality as well as other English courses because of my introvert personality.I feel more comfortable interacting with people in virtual environment… (SF4-3) Participant SM2 stated that he would not use LMOOCs for the following reasons: I don't think I will use MOOCs in the future because I think the instructor and other learners ignored my responses all the time.I don't know why but I just logged off and have not used it after a few times of being ignored….Being ignored in the virtual world also hurts… (SM2-2) SM2's counterparts, SF4 and SF5, seemed to agree with him by nodding their heads.To con rm their meaning and understand their thinking, a follow-up question 'what did you do if you were ignored by the instructor or your peers?' was asked.They responded as follows.
[W]ell, I understood that sometimes the instructor or other participants probably did not understand my questions or statement and therefore did not know how to make proper responses so I felt quite all right….nevertheless, I would feel much better if someone could respond to my questions or statements… (SF4-2) I am not sure if I have had experience of being ignored by anyone in the LMOOC but I just don't like that kind of feeling… (SF5-2) Thus, overall, when participants were asked the underlying reasons they intended to continue or not continue learning with LMOOCs, positive reasons included the opportunity to interact with peers of various cultural and educational backgrounds, and also that LMOOCs enable learners with introverted personalities or various learning styles to t in.In terms of reasons not to continue, they noted that LMOOCs lack face-to-face interaction, which they considered the essence of language learning, and suggested that this would be the greatest downside for participants.Some of them also said that being ignored by instructors or peers was a reason for them not to continue.After considering the pros and cons of LMOOCs, nine of these fteen participants wished to continue using LMOOCs.
RQ3: What are the biggest bene ts and challenges participants have encountered in their experience using LMOOCs in hospitality English learning?
Most participants talked about bene ts in the rst two focus interviews but gradually also started to discuss the challenges they experienced in the last interview.For example, participants SF5 and NF4 shed light on the bene ts of LMOOCs in the rst interview: [T]he LMOOC was fantastic because I can retrieve any information about the subject matter anytime anywhere because they are just there and they never disappear as long as I have internet access….(SF5-1) [T]o me, the greatest bene t about LMOOCs was that I can get to observe others' interactions and learn something out of them.Other than this, I personally consider myself a slow learner and I have to take time to digest newly acquired information and I could take time with LMOOCs… (NF4-1) In the latter interviews, these two participants talked about challenges more than bene ts, and intriguingly, the bene ts they had previously mentioned turned out to involve strong challenges as well: I am still fond of the convenience of LMOOCs but somehow I felt it challenging for me to use it…the problem is not LMOOCs but me.I got lazy way too easy and without [an] external driver… (SF5-3) [A]s I mentioned before, I liked to learn new things through observing others rst but the thing is that not so many participants [came to] interact via LMOOC gradually and therefore I had no one to observe….(NF4-2) The bene t of LMOOCs brought up by the participants in this round of interviews was that they were able to learn by observing others' interactions and to disregard the affective lter that they might have in traditional classes.On the other hand, the challenges they faced was that they had too much latitude with LMOOC; unless learners had extremely strong motivation, they would gradually drop out.After the qualitative data were analysed, a conceptual model of the applicability of MOOCs in Hospitality English based on the framework analysis (see Figure 7) was developed: As shown in the model, the bene ts of this LMOOC described by the participants can be grouped into those related to exibility and to convenience, both aspects of the LMOOC's functionality.The concept of 'affordance' refers to the relationship between user and object (e.g., here, a particular device or virtual environment) (Conole & Dyke, 2004).Given that this study has revealed downsides of LMOOCs including lack of face-to-face interaction and being ignored by peers or instructor, which may be easier to overcome in traditional classrooms through contextual cues or body language, the affordances of LMOOCs can only be fully achieved when such environmental factors are taken into account along with functionality.

Discussion, Limitations And Implications
The present study was designed inductively to bring forth insightful information about the applicability of LMOOCs.This study adopted grounded theory to interpret the data and identify issues faced by this underexplored learning platform.The results of the data analyses elicited some ndings that deserve further discussions.
The ndings of this study rstly shed light on participants' viewpoints on their learning experience with LMOOCs and both upsides as well as downsides have been identi ed.Over half of the participants would continue to use LMOOCs for English language learning.This study further explore the reasons why they would or would not continue to use LMOOCs.Subsequently, this study went on to investigate the bene ts and challenges participants have encountered in their experience of using LMOOCs to learn English language.
The bene ts participants mentioned in the interviews included the exibility and convenience of LMOOCs which echoed the statement of Schulze and Scholz (2018).Flexibility involved 'making it possible to observe others', ' tting learners' various personalities', and 'allowing interaction with different learners', items that re ect the theory of social learning (Clerehan et al., 2003;Miller & Lu, 2003;Samson, 2015); in contrast, convenience items 'allowing retrieval of materials' and 'helping learners learn and acquire knowledge' are advantages of computer-assisted learning.
Most participants agreed that the functionality of LMOOCs enabled them to learn under greatly reduced limitations of time and space, in line with Ryan et al. (2016).Such advantages may be only partly perceived by learners of today, who are considered 'digital natives' and have grown up with the technological savvy for daily application of technical tools, and may thus take them for granted (Prensky, 2001).In this context, the ability of LMOOCs to disseminate knowledge via a virtual environment can be deemed a natural evolution of education.As Gaver (1991) explains, the 'affordances' of a speci c technology are its inherent and important properties, and the aforementioned functionality transported by the virtual environment can thus be viewed as the key affordance of MOOCs (Kreijns, Kirschner, & Jochems, 2002).
Alongside this affordance of LMOOCs, some downsides were also expressed by participants.For example, their interest in learning and/or motivation to learn with LMOOCs seemed to wane after few sessions.In this case, lack of the intimacy found in face-to-face interaction was the main reason for the participants' demotivation; thus, MOOC designers would be well advised to include more collaborative activities (Lin, Cantoni, & Murphy, 2018) or combine pure LMOOCs with traditional classroom teaching as suggested by Wang, An, and Wright (2018) because it still seems to be challenging to create adequate socialization in the virtual environment (Schilze & Scholz, 2018) However, when planning collaborative activities in MOOCs, participants' willingness to partake is another issue for discussion.In this study, some participants said they preferred to observe others on MOOCs but not actively engage, which echoes a rationale of the theory of social learning (Goodyear, 2005).This preference stemmed from their personality traits or learning styles; they did not elaborate whether social identity was a barrier, but did assert that being ignored by the instructor or peers and it remains to be clari ed whether being ignored in this way was a result of their social identity?This study thus suggests that it is vital for MOOC instructors wishing to foster interactions to make sure that no students are isolated, because being ignored by others will activate the pain matrix of the individual's brain (Eisenberger & Lieberman, 2004;Restak, 2009).
Another issue raised was whether observation itself can be considered to be a form of engagement in social interaction, in that learners will place themselves in the 'silent period' when building up their own knowledge through listening and observation (Martín-Monje, Castrillo, & Mañana-Rodríguez, 2018;VanPatten & Benati, 2010).Future research on learners' engagement in LMOOCs may also take this issue into consideration.
At the same time, the effectiveness of learning through observing others on MOOCs is worthy of efforts for further exploration.A study on the adoption of Internet and memory indicated that when learners realize that the information they need is always there on the Internet and they do not need to dedicate effort to remembering it, the cognitive consequences are drastic (Sparrow, Liu, & Wegner, 2011).Some participants in the present study stated that one of the bene ts of learning with MOOCs happened to be the affordance that learning materials are always available on the platform.More empirical evidence will be needed to examine whether such an affordance will compensate for the abovementioned cognitive alteration.
In light of the present research, we call for such a shift globally and a suggestion that educational institutions take on their social responsibility and collaborate with industry to provide further training for in-service employees; in this way, the cost of training will be reduced and the skills of hospitality professionals will improve (Ryan et al., 2016).This type of collaboration will thus generate mutual bene ts for universities and industry; however, the high dropout rate of MOOCs inevitably makes the effectiveness of such courses questionable, and hence their applicability remains debatable.Stracke (2017) proposed a new perspective on this high dropout rate, suggesting that it re ects participants' diverse goals in taking this course; that is, MOOC dropouts are at least in part those who have achieved their learning goals in the course and do not need to stick around, not just the struggling or demotivated.
Furthermore, a study conducted a decade ago by Woody, Daniel, and Baker (2010) indicated that students, regardless of their prior experience with computers and related technology, preferred printed textbooks over the electronic versions.The development of education technology has changed drastically; hence such a statement may not re ect the views of all students today.Nevertheless, traditional ways of teaching and learning still have something that cannot be replaced by technology-perhaps blended education, which integrates both traditional instruction and technology, is more suitable for language learning (Karaaslan & Kılıç, 2019).Last but not least, the affordance of MOOCs may also cause some unexpected outcomes for learners; for example, MOOCs lead to laziness on the part of learners because they grant learners too much latitude in pursuing student outcomes and completing assignments, lowering the motivation for learners to nish the course.Moreover, some participants of this study also pointed out that their lifestyles were conditioned by the MOOCs and their cognitive abilities were also reshaped by the MOOCs.from the psychological and neuroscience perspectives, to provide us with more diverse examinations of the applicability of MOOCs in hospitality English.
Empirical evidence of recent studies (Du & Wang, 2019;Karaaslan & Kılıç, 2019;Pikhart & Klimova, 2019) shows that a blended model that makes the most of both online and face-to-face learning can complement each other and optimize the applicability of LMOOCs.
There are some limitations of this present study.The rst limitation is about the variables of participants' characteristics as workable people and course as an ESP are not fully explored in this study.Many other dimensions, such as learners', instructors, course, technology, design, and environment would affect students' points of view on LMOOCs.Therefore, this study cannot cover as many factors as possible and future studies are advised to include more factors for further exploration.Another limitation is about the generalizability of the results elicited from this study.As a qualitative study, sample size is usually small and caution is advised when the results of this study are to be generalized to participants of various backgrounds.

Conclusion
Given that MOOCs have developed exponentially since 2012, they have drawn much attention in academia.This present research was undertaken with the aim of exploring the applicability of LMOOCs in Hospitality English.Fifteen participants went through three rounds of focus group interviews, and insightful information was extracted from saturated data.Two functions, exibility and convenience, respectively re ecting representing the realms of social learning and computer-assisted learning, were inductively derived from participants' opinions through framework analysis.However, participants also revealed some negatives to the use of LMOOCs in hospitality English learning.Their statements about being ignored by peers and the instructor as well as the lack of the warmth of real face-to-face interaction are related to environmental factors affecting successful implementation of web-based learning.This does not imply that these problems do not exist in traditional classrooms; nevertheless, face-to-face instructors may use contextual cues or body language to address and ameliorate such issues.If LMOOC practitioners can come up with approaches or strategies to deal with such environmental factors, the affordances of LMOOCs will be fully activated.
Two other caveats should also be taken into account.First, the affordance of MOOCs does not require learners to memorize a great deal of information, as it is all stored online; therefore, learners' memorization and retention capacity may be changed by MOOCs, and they may also experience demotivation related to lack of face-to-face interactions.These issues will need more empirical evidence, which can build on the approaches and ndings of the present study.
This study has made the following contributions; rstly, it has explored the applicability of MOOCs in hospitality English through the lens of user(s') experience(s) and their intention to use MOOCs in future education.On this basis, pros and cons of using MOOCs in hospitality courses were noted.Second, based on the data, a conceptual model of the applicability of MOOCs in hospitality English was presented, incorporating important elements such as exibility, convenience, functionality, and affordance.Finally, the rigorous design of this qualitative research depicts the phenomenon of EFL learners' viewpoint about LMOOCs with insightful information and enhance our understanding about this phenomenon as Williams and Moser (2019) stated.

Declarations Figures
Page 17/20 5. Construction of a new network between the new codes and the speci c codes of EFL learners' use of LMOOCs, .Mark the excerpts from the texts, referring to the transcriptions of the interviews with the new codes, 7. Identi cation of common factors among codes, .De nition of the substantive theory of EFL learners' viewpoints on LMOOCs, including the advantages and disadvantages.

Table 1 -
Demographic Information of Focus GroupParticipants Region Gender Code Age Time (Hours) of using LMOOCs Frequencies of login LMOOCs