Pre-Service Teachers’ Perceptions on Integrating a Social Networking Site for Language Learning into English Classes*

This paper aims to report on partial findings of a research study intended to explore how Social Networking Sites for Language Learning (SNSLL) can be integrated into English classes of a BA program in Teaching of Modern Languages. More specifically, this paper reports on the students’ perceptions before and during the implementation of a SNSLL (Busuu). The data was gathered through two collection techniques: surveys and learning logs. Regarding the participants’ perceptions before the use of Busuu, the data analysis shows that, despite some concerns about communicative ability and personal privacy, they perceived this kind of virtual communities as effective platforms to learn EFL through genuine interaction with other users. Although the students’ perceptions during the use of Busuu involved some usability limitations and technical issues, they reaffirmed the potential of this SNSLL to not only facilitate language learning, but also foster intercultural development and metacognitive processes.


Introduction
Currenlty, one of the main concerns in many language teaching contexts is how to provide EFL students, especially those who aim to become language teachers, with a variety of meaningful language learning opportunities that extend beyond the classroom. It has been broadly argued that new technologies have the potential to uncover different ways students could experience language learning meaningfully (O'Dowd, 2007). With this trend towards using new technologies for teaching and learning purposes, social networking sites have become increasingly important because of their potential to bring consistent interaction opportunities that can enhance users' language learning (Lomicka & Lord, 2009). In this context, this report provides partial findings of a research study intended to explore how Social Networking Sites for Language Learning (SNSLLs), such as Busuu, Livemocha, Lingq, Babbel or Duolingo, can be articulated within face-to-face English classes of a BA program in Teaching of Modern Languages at a private Colombian university (Bogotá, Colombia). This paper, however, reports specifically on the pre-service teachers' perceptions before and during the implementation of the SNSLL (Busuu).
Social media has shown potential to improve foreign/second language pedagogical practices, as discussed by different scholars who examine the connection between L2 teaching and social network sites (SNS) such as Facebook, Instagram, and Twitter (Lomicka & Lord, 2009;Mills, 2011;Solmaz, 2017). A more organic way to articulate social media and language learning has been the development of language learning websites with social network features or Social Networking Sites for Language Learning (SNSLL). Although SNSLL emerged in 2007 when Livemocha was launched, as a field of inquiry it is still in its infancy. As Liu et al. (2015) assert, "there is a lack of understanding of how SNSLL can be used to facilitate teaching and learning for English as Second Language (ESL) instructors and students" (p. 113). Additionally, despite the fact that some research has been conducted in the field (Álvarez Valencia, 2016a(Álvarez Valencia, , 2016b(Álvarez Valencia, , 2016dClark & Gruba, 2010;Liaw, 2011;Zourou, 2013), the explorations appear scant in the light of multifarious phenomena that emerge in telecollaborative intercultural encounters (Zourou, 2013).
One context of SNSLL implementation that has suffered from a dearth of inquiry is teaching education programs. At the time of the writing of this paper, except for Álvarez Valencia and Fernández (2019), no research was found that focused on the impact of using SNSLLs in teaching training programs. Thus, the goal of this study was to determine how the SNSLL Busuu can be integrated into face-to-face classes to help pre-service teachers develop communicative capabilities. This paper focuses on preliminary results from two of the data collection tools implemented in two language classes of a teaching education program. The findings report on the pre-service teachers' perceptions about the use of the aforementioned SNSLL.
This study is especially relevant to the language teacher education field since, as stated by Orsini-Jones et al. (2013), SNSLLs have the potential to allow learners to become teachers as well. This blurry line of users as learners and teachers is possible because SNSLLs promote a collaborative-learning environment based on rewards systems to encourage users to help others to learn the target language (Álvarez Valencia, 2016b(Álvarez Valencia, , 2016d. Regarding how SNSLLs are perceived, diverse views have been put forth. Some scholars (Brick, 2011;Jee & Park, 2009;Liaw, 2011;Lin et al., 2016;Liu et al., 2015) point out that the opportunities to interact with native speakers (of the target language) and to receive feedback from them have been positively perceived by users. Brick (2011) and Lin et al. (2016) also found that some users of Livemocha had very positive perceptions about the learning experience in general. Nevertheless, some scholars claim that there are some negative perceptions as well.
Brick (2011), Jee and Park (2009) and Liaw (2011) report that the pedagogical quality and usability of some materials and activities in SNSLLs are rather mechanical, and that their interfaces are somehow difficult to navigate. More recently, Liu et al. (2015) conducted a usability study comparing four SNSLLs (LingQ,italki,and Polyglotclub). The authors report that to provide a good learning environment and learning experience, developers should consider the relevance of three design aspects: information design, interface design, and interaction design.
One of the most frequently commented perceptions on SNSLLs is that they foster motivation (Álvarez Valencia, 2016d;Brick, 2012;Clark & Gruba, 2010;Fan, 2015;Lin & Warschauer, 2011;Lin et al., 2016;Liu et al., 2015;Stevenson & Liu, 2010). Users frequently state that SNSLLs increase motivation and enjoyment due to opportunities they have to collaborate and perform real communicative interactions with other speakers (including native speakers) that, as stated by Thorne et al. (2009), rarely occurs in traditional language classrooms. Interestingly, in their autoethnographic work attempting to learn Korean in Livemocha, found that Clark's experience was shaped by a path of motivation, frustration and demotivation. According to these researchers, motivation was especially triggered by the interaction with other speakers, the peer feedback system, the possibility to meet other users, the sense of progression and the idea of belonging to a community. Clark and Gruba (2010) argue that frustration took place due to audio-lingual-oriented tasks, decontextualized grammar, and vocabulary, the fixed presentation of lessons and lack of adequate support. These scholars also found that demotivation was especially influenced by performing very repetitive tasks, extended grammar drills, level of difficulty, and excessive work when providing feedback to other users. Other researchers, such as Orsini-Jones et al. (2013) have also found that when using SNSLL such as Busuu and Livemocha, over the time the initial motivation in some learners can turn into frustration due to difficulties when interacting with other users and to misleading labeling of language levels.
In line with Clark and Gruba's findings, other researchers such as Potolia et al. (2011) have pointed out that SNSLLs (particularly Babbel, Busuu and Livemocha) may reproduce traditional perspectives and roles of learning. This is due to the structural nature of many of the grammar and vocabulary tasks found in these virtual learning environments that evoke elements from traditional teaching approaches such as the Audiolingual Method and the Lexical Approach (Álvarez Valencia, 2016b). Despite this criticism, Fan (2015), Liu et al. (2015) and Pélissier and Qotb (2012) find that these learning communities have potential to complement language learning in formal learning contexts.
Regarding whether SNSLLs have the potential to foster the development of communicative skills, there are some inconclusive results as well. Based on surveys, screen captures, learning journals and interviews, Fan (2015), Lin et al. (2016) and Chotel (2012Chotel ( , 2013, for example, state that the potential of these virtual spaces to develop communicative skills is limited; nonetheless, the participants in these studies reported that they had actually improved their language skills thanks to the SNSLL used (Livemocha and Busuu, respectively). Harrison and Thomas (2009) state that Livemocha not only facilitates learning progress but, more importantly, it brings about "new modes of active learning" (p. 121). Research has shown that SNSLLs have the potential to foster intercultural communicative competence. In examining students' development of intercultural competence, Álvarez Valencia and Fernández (2019) found out that Livemocha increases students' opportunities to develop attitudes of curiosity and openness, cultural knowledge, and skills for interpreting and getting on well with people of other cultural groups.
From an empirical and theoretical perspective, Álvarez Valencia explored the underlying views of language (2016b), views of learning (2016d) and identity construction (2016a) on Busuu. By drawing on an autoethnographic study of his experience as a language user of the website, and a multimodal social semiotic analysis of the structure and content of the SNSLL, the author concludes that the website is an ecological space where different discourses about language and learning converge in conflicting and complementary ways. Through their semiotic design (e.g., content, iconography, navigational pathways), SNSLLs and, in general, language learning websites construct and position their users in various identity positions that members of the website accept or contend through their use and interactions in the digital environment (Álvarez Valencia, 2016a(Álvarez Valencia, , 2017.

Informal learning vs. formal learning: The potential of SNSLLs
Several scholars have pinpointed the existing gaps in studies about SNSLLs; one of them is the possibility to account for what learners can do in these digital participatory environments, while a second gap concerns with the extent to which these environments can be articulated with formal language classes, different teaching-learning contexts and educational levels (Álvarez Valencia, 2016a(Álvarez Valencia, , 2016cPélissier & Qotb, 2012;Liu et al., 2015). Coping with these two gaps, however, does pose some challenges at both epistemological and curricular levels. Vavuola (2007), for example, has argued that "moving away from 'fixed', traditional classroom learning into more diffused, informal, mobile situations, the learning context becomes vaguer and harder to establish and document for the researcher" (p. 8). From a curricular perspective, one main challenge has to do with the new social dynamics that SNSLLs have created. Informal Web 2.0 language learning activity is crisscrossed by several variables that have changed the current semiotic landscape and traditional textual habitats including the multimodal turn of communication and phenomena of deterritorialization, depersonalization, and detemporalization (Álvarez Valencia, 2016c(Álvarez Valencia, , 2018 of education.
One of the main features of the social web and SNSLL is their multimodal nature. Particularly, SNSLLs are interfaces that combine different modes of communication for meaning-making.
Thus, linguistic, visual, aural, spatial and interactive modes combine to generate a multisensory learning experience. Likewise, multimodality calls into question the dominant verbocentric perspective in TESOL, where the concept of communication seems to have been reduced to developing competency in the written and verbal modes of communication, at the expense of other modalities such as gesture, gaze, spatiality among other that part-take in interaction and meaning-making (Álvarez Valencia, 2016c(Álvarez Valencia, , 2018. The collapsing of contexts (spatial, social, and temporal boundaries (Boyd, 2009) brought about by SNSLLs challenges traditional educational processes because learning a language is deterritorialized, that is, the institutionalized language classroom ceases to be the main locus of input; the act of teaching is depersonalized, which means that learners do not depend on the authoritative figure of a teacher; and learning and socialization are detemporalized -access to contents is not bound to any institutional schedule. The challenges yielded by SNSLLs and their potential to be integrated in formal contexts of language teaching bring to the fore the discussion about the promises of informal learning in contrast to formal learning (Schugurensky, 2000;Toffoli & Sockett, 2010).
Formal learning refers to the learning process that is bound to an educational institution and that emerges from pedagogical tasks provided in the classroom or as a result thereof (e.g., homework). By contrast, informal learning is defined as "any activity involving the pursuit of understanding, knowledge or skill which occurs outside the curricula of educational institutions, or the courses or workshops offered by educational or social agencies" (Livingston, 1999, p. 51). SNSLLs are good examples of spaces where informal learning materializes because it occurs outside the curriculum of formal and non-formal educational institutions. In fact, as discussed by Toffoli and Sockett (2010) evoking Cross (2006), informal learning provides learners with unofficial, unscheduled, impromptu learning experiences. Applied to the context of SNSLLs as informal learning sites, we find that the learning material offered on the website is not prescribed by a teacher or makes part of an official language class. Drawing on the concept of User Generated Content (UGC), users are generally free to design, disseminate, and manipulate their own or others' content. Following the idea that content is not prescribed is the condition that learners do not have to follow a pre-established schedule and thus can construct their own learning trajectories. This, in turn, opens up the doors for spontaneous engagement with learning materials or telecollaborative interaction with other learners. Unlike formal learning, in user-initiated social networking spaces learners can take different learning trajectories and look for different language peers from different cultural and linguistic backgrounds without spatial or temporal limitations.
Although SNSLLs indeed offer multiple affordances for language learning and intercultural negotiation, the question remains whether these digital environments could be harnessed to enrich language learning in formal educational contexts. The work by Álvarez Valencia and Fernández (2019) and the research reported herein constitute an attempt either to make formal language learning more informal or to make informal language learning more formal, with the ultimate purpose of enriching the learning experience of future language teachers.

Social network sites (SNSs) vs. social networking sites for language learning (SNSLLs)
The use of the word 'network' and 'networking' points at one of the main features that distinguishes digital social spaces such as Facebook (SNS) from digital educational spaces such as Busuu (SNSLL). Traditionally, SNSs like Facebook or Ning have been used to maintain contact with acquaintances, friends or relatives with whom people already hold a relationship offline in spaces such as family, work, neighborhood, or interest groups (Haythornthwaite, 2005;Zourou, 2013). Thus, the main purpose of an SNS is to maintain a users' network of individuals rather than engaging in the act of networking. On the other hand, there are digital social spaces with educational purposes whose main interest is to connect people (networking) who, in most cases, do not have pre-established connections offline (Boyd & Ellison, 2007).
This group of digital spaces that combines features of social media with educational purposes can be called educational social networking sites (ESNS). ESNS's encompass digital teaching and learning environments about one or various areas of knowledge like math, social studies or foreign languages. Within this broader category, we can locate Social Networking Sites for Language Learning (SNSLL). SNSLLs, in turn, belong to the broader category of language learning websites (LLW) that, according to Álvarez Valencia (2017), have developed along with the technological affordances of Web 1.0 and Web 2.0. SNSLLs represent the latest generation of LLWs that draw on principles of interactivity, multimodality, and community.
SNSLLs like Busuu connect people who want to learn a language by facilitating interaction with native speakers or other learners to help each other in their learning process. These virtual communities offer different language courses organized in levels of proficiency, free access to materials (or wider access through a paid subscription), and interactive activities including pronunciation, dialogues, listening comprehension, vocabulary, translation, peer correction, video chat, and a reward system (Liaw, 2011).

Research design
This study follows the principles of a mixed-methods approach (Creswell & Plano, 2010) because it combines procedures, data collection instruments, and data analysis techniques from qualitative and quantitative approaches. In the first place, the qualitative nature of the present study is related to a holistic, interpretative, and naturalistic perspective (Marshall & Rossman, 2011;Patton, 2002) to examine the participants' learning processes when they become Busuu users within the context of an undergraduate teacher training program. In the second place, the study involves quantitative procedures and tools such as the use of descriptive statistics. Hence, the selection of a mixed approach aimed at facilitating the verification of findings and the validity of the study (Creswell & Plano, 2010).
As a part of a larger research project that aimed to explore the integration of two SNSLLs (Livemocha and Busuu) in the language programs of two Colombian universities, this paper focuses on the initial results of implementation of Busuu in two courses in one university. The courses belonged to the B.A. in Teaching of Modern Languages in a private university, whose English program comprises seven levels. The participants of the study were enrolled in the second and fourth levels, Basic English (Initial stages of B1, based on the CEFR 1 ) and Low-Intermediate English (Initial stages of B2, based on the CEFR), respectively. The participants (n = 36) were 19 the pre-service teachers (6 men, 13 women) enrolled in the Low Intermediate class and 17 the pre-service teachers taking the Basic English class (6 men, 11 women). The ages ranged between 18 and 22 years old. In the results section, the participants are referred to as P1 (Participant 1), P2 (Participant 2), and so on.
The information was gathered during an academic semester through a survey and students' logs.
The survey, administered before the pedagogical intervention with the SNSLL, aimed to determine the students' demographic information and their previous knowledge/experience with SNSs or SNSLLs. The students' logs consisted of a set of closed and open-ended questions that students answered after every working session on Busuu. The log sought to identify the preservice teachers' perceptions, activities, and issues that they faced during their work on the SNSLL. In total 264 logs were collected during the data collection period which lasted an academic semester (18 weeks).
Before the implementation of the SNSLL in the English lessons, the researchers, who at the same time acted as instructors of the two language classes, aligned the contents of Busuu with the syllabus of each class. The English lessons were delivered daily (a 2-hour session every day from Monday to Friday) and two weekly sessions were devoted to the integration of the website with the classroom activities. The participants were also asked to carry out independent work on Busuu outside class time. In addition, they were asked to use the chat tool to engage in interaction with community members as well as with their own classmates. Some of the activities developed on the website were formally assessed. Initially, the classroom sessions with Busuu were structured following this pattern: 1) Warm-up or introductory activity, 2) work on Busuu's lessons, which could be completed in any order, 3) written or video chat with other users or classmates, and 4) Log entry completion. In some lessons, students were instructed to write compositions so that other Busuu users could provide feedback; some other times, time was used to review other users' writings. As the semester progressed, the sequence of development of lessons became more flexible, varied, and mixed with other kinds of materials and activities; for example, online videos and games.
To analyze data, the researchers looked for patterns within each instrument and across the two instruments: survey and students' logs. Following Marshall and Rossman's (2011) procedures, the data items of qualitative nature from the survey were collated, then coded searching for patterns and themes. Questions of a more quantitative nature were analyzed in terms of frequencies and average percentages. A similar process was conducted on the students' logs, however, questions not directly related with the purpose of this initial report were not considered.

Results and discussion
As stated previously, the present study sought to report on the students' perceptions before and during the pedagogical intervention using the SNSLL Busuu. The first part of this section reports on students' views about SNSLLs before using one and the second part focuses on students' perceptions about the experience of learning English on Busuu.

Students' perceptions before the implementation of Busuu
In order to account for the learners' perceptions (n=36) about SNSLLs before implementing them, a survey was administered. The survey was structured so that the participants could report their perceptions about 1) Information and Communication Technologies (ICT) in general, 2) Social Networks/Networking Sites they had used to learn languages, 3) SNSLL vs traditional resources for language learning and 4) their expectations and concerns about using SNSLLs in the English class. In the interest of space, we will only focus on the last three sections of the survey because they relate directly to the main topic of the study.
[INSERTAR FIGURA 1] One of the questions included in the survey inquired if students had used SNSs to learn languages. Youtube topped the list, chosen by two thirds of the population (63%), followed by Duolingo with 50%, and Facebook with 42% (figure 1). These results evidence that the participants, as well as other language learners around the globe, are increasingly accessing educational scenarios that transcend national territories, formal institutions and fixed teachinglearning schedules (Álvarez Valencia, 2016c;2018) mediated by these virtual spaces. These results support the need for more research in the SLA field (Clark & Gruba, 2010;Liu et al., 2015), not only due to the pedagogical potential of SNSLL, but also to the growing number of language learners who seek complementary or alternative ways to develop L2 competency through these digital environments.

Figure 2. Perceptions on why students used websites for language learning
Source: own elaboration Participants were then asked why they used social media (not necessarily social networking sites, for language learning). Responses show that 82% used them to reinforce vocabulary, 74% to practice listening comprehension, and over 60% replied that SNSs were used to practice grammar and improve pronunciation. These results suggest that many participants adhered to a structural view of language learning, which probably led them to believe that the key factor to learn a language lies in mastering vocabulary and grammar. On the other hand, those who have used websites to interact with others (37%), learn about a foreign culture (13%) and practice speaking (29%) seem to instantiated a communicative view of language where authentic interaction plays a key role to learn a language. No doubt that when teachers draw on these technologies, they shape learners' expectations and beliefs regarding the technology's potential, use, and effect on teaching-learning processes. Interestingly, despite the fact that Álvarez Valencia and Fernández (2019) recently concluded that Livemocha has the potential to help users develop intercultural skills, knowledge or attitudes, only a small percentage of the participants in the present study seem to use websites, including SNSLLs, with intercultural learning goals in mind.
In the survey, participants were also asked to state their opinion about SNSLLs in order to determine how they felt towards them before implementation. Overall, as seen in the following excerpts 2 , they deemed these sites as effective, innovative, and adaptable tools. I think it's an excellent tool since it is not only innovative but also effective and adapts easily to the learner's needs.  (Clark & Gruba, 2010;Lin & Warschauer, 2011;Lin et al., 2016;Liu et al., 2015;Stevenson & Liu, 2010), but also to the "novelty effect" that SNSLLs usually bring about (Brick, 2012). However, these initial positive response may be called into question when it comes to keeping students' engagement for longterm implementation within regular face-to-face language lessons (Brick, 2012). As suggested by Clark and Gruba (2010), besides avoiding overuse, the implementation of such virtual environments in the language classroom demands creativity and variety through other activities that may somehow contribute to keep and even enhance the level of students' motivation.
Despite the generalized positive expectations reported by the students, two of them were not keen on using SNSLLs in the English course, as seen in the excerpts below: they are not effective and time is wasted most of the times […] they can be distractors during the learning process since society is highly involved in communication through social networks, and they wouldn't be 100% focused on learning the language. (p. 26, 2016). I don't think they are actually effective, and often times you just waist you time (p. 9, 2016).
The pre-service teachers also responded to the comparison between SNSLLs and traditional resources for learning (e.g., language textbooks). They privileged the former arguing that SNSLLs were more interactive and engaging, and that learning was more authentic.
When we learn through social networks, it is a more real context. (p. 7, 2016) we interact in a realistic environment and knowledge is applied interactively. (p. 29, 2016) learning in the SNSLL is more fun. (p. 12, 2016).
You can learn [grammar] rules from books, but with SNSLLs you can interact with native speakers (p. 8, 2016) As depicted in the excerpts above, our students supposed that the key differentiator to favor SNSLLs over textbooks (and other traditional resources) lied in that these virtual spaces would allow them to experience genuine communicative encounters, including interaction with native speakers. This generalized perception aligns with what other scholars have found to be the most important potential of using SNSLLs (Clark & Gruba, 2010;Fan, 2015;Lin et al., 2016;Liu et al., 2015). One possible explanation for such a positive expectation regarding the real communication scenarios that SNSLL can offer may correlate with Fan's (2015) findings in her study with a college-level English course. Her students reported that "using this platform [Busuu] did decrease the mental distance between them and their international Busuu pals. The compressed psychological space created via cyberspace increased the frequency, spontaneity, and convenience of on-line language and culture exchanges" (p. 62). In this line of thinking, our own participants were probably predicting not only that they would value the idea of engaging in authentic intercultural exchange, but also that they could embrace the act of performing these online interactions that, when compared to face-to-face communication, may ease L2 anxiety and decrease cultural distance.
One participant, however, preferred textbooks since content "in the SNSLLs is much more informal while one is certain of the good quality in the books" (p. 31, 2016). Despite being just a perception before Busuu was actually brought to our English class, this student's concern somehow resonates with what some SNSLL studies have indicated, that the pedagogical quality of some of the contents and activities of these online-educational environments is questionable in terms of activity design or underlying theories of language and learning Brick, 2011;Jee & Park, 2009;Liaw, 2011;Liu et al., 2015;Orsini-Jones et al., 2013).
With regards to the participants' expectations about integrating the SNSLL into the English class, they predicted positive experiences as well as learning outcomes, as the excerpts below hint at: with the integration of SNSLL, the youngsters can improve their communicative competences and develop new skills in the real interaction with the language. (p. 17, 2016).
[my expectation relates to] using the language in a real context, I mean the natural use of the language; also, the interaction with native speakers and their feedback. (p. 34, 2016). receiving feedback from other people is one of the most interesting aspects of this SNSLL since it allows us to interact with others. (p. 31, 2016).
I think classes will be more dynamic and different; they won't be traditional. (p. 25, 2016).
These perceptions were particularly relevant to the present research project because they suggested that our pre-service teachers, in general, excitedly welcomed the idea of articulating the potentials of SNSLL with common classroom activities. Although we do not uphold nativespeakerism, student's discourses indicate that the native English speaker is still considered the model to guide language teaching and learning. Students hoped for communicative improvement in the target language, authentic interaction with native speakers of Englishincluding their feedback-and more stimulating learning opportunities. Interestingly, regarding the feedback system of SNSLLs, various scholars have concluded that such a feature is not usually reliable and it can even become cumbersome for learners (Liu et al., 2015;Orsini-Jones et al., 2013). In their study of Bussu with undergraduate and graduate TEFL students, Orsini-Jones et al. (2013) found that feedback lacked reliability due to its inconsistency and focus on grammar and vocabulary, overlooking other dimensions such as context and semantics.
On the other hand, as evidenced in the excerpts below, the participants had different concerns, including the challenges they would face in communicating with other SNSLL users: [I'm concern] that the other people I will talk to might have a different purpose to interact with me (p. 4, 2016).
Not being able to understand or to make myself understood when it comes to interacting with other people (p. 22, 2016).
My biggest fear would be realizing that my knowledge is not really deep in terms of technology and language proficiency (p. 18, 2016).
A possible fear that all this may bring about is that the use of these social networks may not be really effective in the process of learning a language or that they may not be really helpful to the development of our class (p. 9, 2016).
Overall, the concerns highlighted by our students reflect what other researchers have shown.
For example, Fan (2015) found that her students experienced a similar fear associated to the quality of their communicative skills when interacting with other users of Busuu; more specifically, her students expressed that, when interacting online with other users of Busuu, grammatical correctness was a salient source of anxiety. P4's comment about privacy and safety issues is not far from reality. Language learners who have taken part in similar studies (Harrison & Thomas, 2009;Lloyd, 2012;Malerba, 2015;Orsini-Jones et al., 2013;Stevenson & Liu, 2010) have even stated that they have experienced some kind of cyberflirting.

Students' perceptions during the implementation of Busuu
Students were asked to fill in a learning log entry after every session they worked on Busuu, From these logs, four emerging themes were identified: language learning, intercultural development, metacognition processes, and difficulties using the interface.

Language learning
Overall, our students reported positive learning experiences, which evokes results from previous studies regarding perceptions about language learning on SNSLLs (Álvarez Valencia & Fernández, 2019;Clark & Gruba, 2010;Fan, 2015;Harrison & Thomas, 2009;Lin et al., 2016;Stevenson & Liu, 2010;Zourou & Loiseau, 2013). As observed in the excerpts below, more specifically, our students claimed to have learned about vocabulary, grammar and culture. They also improved writing and pronunciation skills.
I learnt vocabulary of different topics; also I could know more about different cultures. (p. 23, 2016).
What I learned the most was vocabulary and some tips to improve writing. Likewise, [I learned] from interacting with people from other countries (p. 6, 2016).
I learnt a lot of vocabulary, cultural elements and some expressions […] you can improve your grammar. (p. 28, 2016).
These results support what has been found by other related studies. For example, Liu et al. (2015), when examining similar SNSLLs to Busuu (LingQ,Italki,and Polyglotclub) concluded that the "features available through these sites revealed new and different ways for learners to improve their knowledge of L2 vocabulary, grammar, and culture" (p. 36). Lin et al. (2016) also claim that their participants, using Livemocha, reported high levels of perceived learning, especially concerning listening, speaking, writing, reading and vocabulary.
Concerning the cultural learning that our students perceived, other studies such as the ones carried out by some scholars (Álvarez Valencia, 2016b;Álvarez Valencia & Fernández, 2019;Lloyd, 2012), also hint at the idea that SNSLLs have the potential to provide language learners with scenarios to develop intercultural knowledge and skills.
The data analysis suggests that perceptions concerning linguistic and communicative progress are consistent with the initial expectations the students revealed before the implementation of Busuu in the English language lessons. There seems to be a slight difference though, the expectations were largely stated through expressions that implied communicative competence or skills, while after having used Busuu, students reported improvement in more specific areas such as vocabulary, grammar and writing. This distinction is relevant because knowledge of the levels of a language do not always translate into communicative capability.

Intercultural development
Among the questions -in the learning logs-that the students were asked to answer after each session on Busuu, there was one that inquired about the (inter)cultural aspects that they could identify during their communicative exchanges on the chat. Some of the themes that emerged from the data sets, regarding discussions about cultural similarities and differences, were mainly related to the contents of the Busuu units, such as fashion, media, arts and politics, as seen in the excerpts below.
We talked about elections in Bogota and how elections function in his country. (p. 25, 2016) We discussed how the media can connect people and make them talk about it (p. 26, 2016) [I realized that] we all are suffering the same problems and we need to get closer and worry more about the world […] I think that specially in this topic -Fashionyou can exchange a lot of important information and learn more about different cultures. (p. 36, 2016) We talked about art, artists and their creations (p. 23, 2016) However, other discussions also took place from a more personal perspective. The main topics during these more personal conversations were related to language learning experiences, leisure time activities, their countries and cities of origin, as evidenced in the following excerpts: We talked about our hobbies and things that we usually do. We talked about important decisions that we have made and we talked about our beliefs. (p. 25, 2016).
We asked each other about our cities and our countries, and […] we talk about our lives. (p. 4, 2016).
We talked about the amount of time he had been learning Spanish, some vocabulary, the reasons/motivations why both of us were learning different languages and some aspects of pronunciation. (p. 33, 2016) The intercultural communicative exchanges portrayed in the excerpts above, along with arguments in other SNSLL studies (Álvarez Valencia, 2016a;Álvarez Valencia & Fernández, 2019;Lloyd, 2012), supports the idea that these virtual educational scenarios have the potential to mediate intercultural learning and foster intercultural awareness due to the authentic communicational practices that take place among their members. However, fostering intercultural competence through SNSLLs should be monitored and accompanied by the language teacher: Since a particular NS [native speaker] of the target culture is not necessarily a representative of every aspect of the culture, leaving the users to solely depend on a certain group of NSs online might lead to a biased perception of the culture or overgeneralization of their acquired cultural knowledge from their online friends (Jee & Park, 2009, p. 455).
on Busuu to the expectations they expressed before they started using the website, it is interesting to note that their current perceptions now seem to transcend the linguistic-and communicative-oriented focus to involve (inter)cultural development as well. This is clearly one of the most relevant affordances of SNSLLs and, particularly, interactive interfaces such as chats that set up contexts where users engage in meaning-making through linguistic and cultural exchange and negotiation. The online intercultural exchanges that SNSLLs such as Busuu foster evokes a wealth of research on the promises of virtual environments that according to O'Dowd (2007) help students to develop intercultural sensitivity and independence as well as understand that communication implies not only the exchange of information but also the expression of identity.

Metacognition processes
The activities developed on Busuu have the potential to facilitate metacognition processes.
Metacognition involves "individuals' awareness and management of their learning processes" (Raoofi et al., 2014, p. 37). While working in Busuu, students were building their own learning trajectories because they could choose the order in which they wanted to develop the lessons of the day. Students' choices drew on their own language needs and interests. In many cases, students were observed taking notes and using other online resources such as Wikipedia, dictionaries or translators to clarify or expand contents from the Busuu lesson. In the logs, students indicated that Busuu allowed them to spot areas of communicative improvement and to monitor their language learning: It allows to get a better knowledge of the language. It helps you realize what skills you need to improve to have an efficient learning. (p. 18, 2016 You think about how you can express to the others in a real situation. (p. 2, 2016) This is indeed a relevant finding since it is well known that metacognition can enhance language learning significantly (Anderson, 2008;Raoofi et al., 2014). It is important to highlight that the participants in this study are pre-service language teachers, who understandably may be more likely to develop metacognitive behaviors when compared to regular L2 learners. More research is needed, though, to further explore how SNSLLs can mediate the development of metacognitive processes.

Difficulties using the interface
The participants also reported their perceptions concerning some difficulties they encountered when interacting with the website tools. The most common struggle was linked to finding a chat partner on Busuu, a shortcoming that was documented by other studies about Busuu and other similar SNSLLs (Álvarez Valencia & Fernández, 2019;Fan, 2015;Liu et al., 2015). Interacting with other Busuu users was a mandatory activity in every lesson (twice a week). However, as portrayed in the passages below, the participants complained about spending a lot of time, sometimes unsuccessfully, finding someone to talk on a video or written chat session.
The main difficulty was finding someone to chat with, many of them [other users] were not available (p. 33, 2016).
I tried to communicate with someone, but nobody replay to me. (p. 2, 2016) If your friend [other Busuu users] doesn't answer, you can't improve. (p. 20, 2016) Another difficulty included technical problems. Since at the time this research project took place a computer room was not available, individual laptops were brought to the classroom, which implied serious and frequent connectivity issues. One participant stated that "the most common problem was the internet connection, the activities kept blocking themselves very frequently" (p. 7, 2016). Both the instructors and the students then had to rely on a Wi-Fi connection that was not strong enough to allow fluid and uninterrupted work in the SNSLL: "There were a lot of problems with the internet connection and loading the website" (p. 11, 2016) The issue of accessibility to appropriate computer equipment and fast Internet connectivity plays an instrumental role when it comes to incorporating these technologies in the language classroom. Furthermore, unlike Liu et al. (2015) study on LingQ, Lang-8, Italki, and Polyglotclub's navigability, our students did not report any operational issues with Bussu, such as "awkward registration process, distracting advertisements, complicated navigations, inaccurate translation, and technical errors" (p. 24).

Conclusions
This paper aimed at investigating the perceptions of 36 pre-service teachers regarding the integration of the SNSLL Busuu into their English classes for over an academic semester.
Regarding students' perceptions before using Busuu, they initially perceived them as effective, innovative and flexible tools that would allow them to engage in authentic and meaningful communication with English native speakers. Furthermore, they also expressed some concerns about potential limitations with their own communicative skills and privacy issues when interacting with other members of the virtual community.
In regard to the participants' perceptions during the implementation of Busuu, the findings were generally positive. Firstly, their initial positive perception about SNSLLs' potential to enhance learning was restated, but with reference to specific areas such as vocabulary, grammar and culture. Interestingly, cultural development was not mentioned by the participants prior to the use of Busuu; however, it became a strong perception during its implementation, supporting the view that these SNSLLs have the potential to mediate the development of intercultural skills and awareness. Secondly, the data sets suggested that the participants regarded Busuu as a tool to boost metacognition processes, which is broadly considered as a fundamental feature for successful L2 learning. Finally, the students reported serious difficulties to find other members of the SNSLL to interact with. They also commented on the technical difficulties that they dealt with due to the lack of proper internet connectivity, which caused some frustration among the participants.
Given the partial nature of this study in that it reports initial findings, one limitation is that it only presents data from two data collection instruments. Nevertheless, we hope that these findings provide insights for both educators and researchers who are interested in understanding students' attitudes towards adapting these digital platforms for language teaching and learning and also how these interfaces may contribute to enhancing language and intercultural development. As we continue with the data analysis of other instruments and other SNSLLs to expand on the initial results presented above, we expect that this study contributes to the insipient research that has explored the potential of SNSLLs in teacher education programs.
Further research should consider exploring the results of using and comparing various SNSLLs in different educational contexts, while also studying the different semiotic repertoires (linguistic, communicative, cultural etc.) that students develop. Like any human artifact, SNSLLs do indeed afford different forms of cognitive, attitudinal and behavioral development, nonetheless it is in the interaction with individuals that the potential of artifacts multiplies. We sustain that the interaction between SNSLLs and language learners is a fruitful area of inquiry for language teacher education.