Journal of Computer Assisted Learning

2.1 Facebook as a learning environment

Facebook is an SNS that allows users to share information, create personal profiles, express ideas on issues that are relevant to them, connect with their family and friends, and build online communities (Facebook, 2018). Note that Facebook was originally intended for nonacademic use. Consequently, several concerns were raised against its use for pedagogical purpose. These include security and privacy issues that may compromise students as they interact with one another and post personal information (Hew & Cheung, 2012). In other cases, some scholars reported that using Facebook for academic purposes distracted students from their academic tasks (Shih, 2011) and limited their level of engagement (Arzu, 2014; Hughes, 2009).

Nonetheless, we should not dismiss Facebook as irrelevant to educational context. It exhibits many features that can be used for academic purpose. These features and the increasing popularity of Facebook continue to excite scholars and educators to explore its educational use (see Aydin, 2012; Manca & Ranieri, 2013; Manca & Ranieri, 2016b; Yang et al., 2011) in either formal or informal context (Manca & Ranieri, 2016b). In fact, several studies reported that it facilitates interaction among learners, provides a venue for peer and teacher feedback, and promotes collaboration and information sharing (Lampe, Ellison, & Steinfield, 2008; Selwyn, 2009; Tapscott & Williams, 2010). It also motivates students (Siegle, 2011), enhances self‐esteem and reduces anxiety (Bowers‐Campbell, 2008; Ellison, Steinfield, & Lampe, 2007; West, Lewis, & Currie, 2009), and facilitates differentiated learning (Hew, 2011).

2.2 Related studies

Currently, only a number of papers reviewed the literature on the educational use of Facebook. For instance, Yang et al. (2011) reviewed 21 empirical research studies on the pedagogical use of Facebook to identify the areas in which research on Facebook has been conducted (i.e., effects on learning outcomes, use of Facebook for libraries, and Facebook usage profile) and the factors that influenced its use in these studies (i.e., trust/privacy, personality factors, and pedagogical issues). Most of these studies were descriptive, conducted in the higher education context, and came from various disciplines such as communication, computer studies, chemistry, psychology, and English language. The paper concluded by calling for more empirical studies on the effects of Facebook on teaching and learning. Similarly, Hew (2011) reviewed 36 articles, which are mostly based on self‐report data. He categorized the studies into three major topics: Facebook usage profile, effects of using Facebook, and attitude towards Facebook. His findings revealed that Facebook has very little educational use and exposes students to privacy issues. Although these two studies presented interesting insights, their findings may not accurately reflect the current landscape of studies on the pedagogical use of Facebook for two reasons. First, the latest reviewed articles were published in 2010. Hundreds of studies on Facebook as a learning environment have been published since 2010. Second, major research databases such as Scopus were excluded. Such exclusion raises question on whether the results provided a complete picture of the phenomenon being investigated.

A similar review was embarked by Aydin (2012) who categorized the studies into six groups: (a) Facebook users; (b) reasons people use Facebook; (c) harmful effects of Facebook; (d) Facebook as an educational environment; (e) effects of Facebook on culture, language, and education; and (f) relationship between Facebook and subject variables. He concluded that Facebook is a viable pedagogical tool as it improves classroom practices and student participation. Although his review provided a picture of the current studies on Facebook as a learning environment, the data source and method used in retrieving and selecting the studies were not clearly articulated. The rationale for the predetermined categories used in classifying the reviewed studies was evidently absent as well. Hence, it is very likely that other relevant studies were wittingly or unwittingly excluded from the analysis.

More recently, Niu (2017) reviewed 57 empirical studies on the use of Facebook for academic purposes. They were categorized into four groups: teaching and learning tool, perception studies, influence on learning outcomes, and student‐initiated use of Facebook. The analysis revealed that majority of the studies were experimental (56%) whereas the rest used survey/observational methods. However, most studies relied on preexperimental design and self‐report data, which make the findings less conclusive. From her review, the data revealed the potentials of Facebook in formal teaching and learning contexts. Other reviews were conducted by Chugh and Ruhi (2018) and Voivonta and Avraamidou (2018) who focused on the educational use of Facebook in higher education setting. Both reviews concluded the pedagogical value of Facebook such as improved performances, student interaction, and engagement. However, they also highlighted the potential problems in using Facebook particularly students' privacy concerns. Although these three reviews showed a current state of Facebook as an educational tool, they lack a clear framework for analysis.

To date, the most comprehensive review on the educational use of Facebook was that of Manca and Ranieri (2013, 2016b) who sought to identify the emerging educational uses of Facebook and how these potentials are translated into practice. Unlike other Facebook reviews, these two papers provided a clear theoretical framework for the analysis. Both of these reviews zeroed in on empirical studies about Facebook as a learning environment. They also investigated the extent the studies on using Facebook as a learning environment exploited the three Facebook affordances, namely, mixing information and learning resources, hybridization of expertise, and widening context of learning. The first critical review included 23 articles (published in June 2012 and earlier), whereas the second paper reviewed 147 articles (published from 2012 to 2015) using a revised classification of categories. All of these articles were generated from journals and major databases, such as Education Resources Information Center, Education Research Complete, Web of Science, and Scopus.

Their first critical review revealed that almost all articles were carried out in the higher education context. Almost half of the articles focused on Facebook as an environment to deliver instruction, whereas others dealt with Facebook as a supplementary or alternative environment and students' reactions to the use of Facebook as an instructional tool. The paper identified five main educational uses of Facebook: (a) supporting discussion and interaction among students, (b) developing multimedia content, (c) sharing resources, (d) delivering content, and (e) supporting self‐directed learning. Their study also revealed that both students and teachers were aware that Facebook allows access to a plethora of resources not available in closed learning systems. Finally, their analysis showed that most of the studies focused on students' satisfaction rather than actual learning. Thus, they suggested that Facebook be treated with caution regarding its educational value.

The 2016 paper of Manca and Ranieri extended the first review by analysing the literature according to three types of educational use of Facebook, that is, formal use in formal learning settings (FUF), informal use in formal learning settings (IUF), and use in informal learning settings (UI). One hundred forty‐seven articles were retrieved using the same search and retrieval strategies used in their 2013 paper. Among these articles, 69 were coded as FUF, 68 were IUF, and the remaining 10 were UI. Most of these studies were conducted in the higher education contexts. In terms of geographical distribution, majority were undertaken in Asian countries followed by North America and Europe. Manca and Ranieri (2016b) concluded that although Facebook is a dynamic and flexible platform for teaching and learning, the three Facebook affordances seemed to be underexplored and partially implemented.

Although there have been critical reviews of literature on Facebook as an educational tool, none have explored a specific field of interest. This study seeks to critically analyse and define the contour of scholarly literature on Facebook as a technology‐enhanced learning environment in language teaching and learning. The study also reflects on some of the epistemological decisions (e.g., context, methodologies used, and foci of interest) that academics have adopted when using Facebook as a language learning environment. Note that this study does not intend to provide recommendations on the use of Facebook in various language learning contexts. Instead, it is oriented at examining the demographics (publication year, source title, document type, geographical area, context of the study, research design, and Facebook uses) and foci of interest of the selected studies as well as the how Facebook was used in these studies.

3.1 Data search and retrieval

I searched and retrieved the selected publications from the Scopus database. Scopus is Elsevier's abstract and citation database that monitors and analyses scholarly literature in various fields, namely, science and technology, mathematics, engineering, health and medicine, social sciences, and arts and humanities (Scopus, 2018). Unlike other databases, Scopus is a more reliable and accurate source of data (Bornmann & Marx, 2014; Jacso, 2008; Meho & Yang, 2007). In fact, other databases (e.g., Google Scholar) were reported to have deficiencies for research evaluation (García‐Pérez, 2010; Jacso, 2009) and inaccuracies (Jacso, 2008). Although Web of Science and EBSCO are equally reliable, Scopus has more comprehensive collections with over 64 million records, 21,500 journals, 131,000 books, and 7.5 million conference papers (Scopus, 2018). Thus, Scopus was chosen by leading independent networks for higher education such as Times Higher Education and Quacquarelli Symonds as a data source for ranking world universities.

The inclusive years for analysis are from 2010 to 2017. After surveying Scopus database, the starting date used was 2010 because no literature on Facebook as a language learning environment has been published yet in any Scopus‐indexed publications before the said year. Published literature (i.e., journal articles, conference papers, reviews, notes, and book chapters) were searched and retrieved using the “Document Search” feature of the Scopus database. To do this, I typed “Facebook” in the search tab and limited the search category to “Keywords.” Because noneducation literature was displayed in the list, the search was further limited to computer science, social sciences, and humanities and arts resulting in 5,221 retrieved documents. The title and abstract of these documents were reviewed to ensure that the selected literature were relevant to language teaching and learning. The “Advanced Search” feature was also used to retrieve additional documents. These codes include KEY (Facebook) AND ALL (language and linguistics), KEY (Facebook) AND ALL (language teaching), KEY (Facebook) AND ALL (ESL), KEY (Facebook) AND ALL (ELT), KEY (Facebook) AND ALL (English language), KEY (Facebook) AND ALL (learning English), and KEY (Facebook) AND ALL (EFL). These keywords were grounded from the relevant words expected to be found in the target studies and the keywords used by studies about Facebook as a learning environment. The code ALL was used to allow the search of relevant words from the title, abstract, keywords, source title, and whole text. Seventy‐one documents were generated using these search and retrieval strategies. These documents were further evaluated to ensure that they satisfy the following criteria: (a) primarily focused on the use of Facebook in language teaching and learning, (b) written in English, and (c) completed research (i.e., those published papers with analysed and reported results). Using these criteria, 30 studies were excluded from the analysis. Twenty (27) of these did not focus on language pedagogy, two were work‐in‐progress during the time of document retrieval, and one was written in Spanish. Thus, the final screening trimmed down the documents to 41.

3.2 Data analysis

After an extensive and principled search of literature, I performed a qualitative review of the 41 documents because they were too nonhomogenous to merit a quantitative meta‐analysis. Nonetheless, a quantitative data analysis was undertaken about the demographics of the selected studies. I identified their publication year, source title, document type (article and conference paper), geographical area, context of the study (primary, secondary, and higher education), research design (quantitative, qualitative, or mixed method), and Facebook uses (FUF, IUF, and UI).

Anchoring on Manca and Ranieri's (2016b) technique, I conducted a qualitative analysis of each of the studies in terms of aims, major findings, and conclusion. I also looked into these studies the usage (e.g., interaction, assessment, discussion, content development, content delivery, and self‐directed learning), features, and reasons for using Facebook. In classifying the studies based on foci of interest, I used multilevel coding (Birks & Mills, 2011). This means that specific categories emerged from a grounded theoretical approach after the analysis of the studies. By using this approach, the categories that emerge from the data were allowed to evolve to accurately capture and classify the selected studies. To do multilevel coding, I identified relevant codes from each of the studies and categorized these codes based on their related properties and dimensions. Then, I performed concurrent and progressive data collection to further shape and enhance the initially identified subcategories. Afterwards, I linked the identified subcategories that are conceptually related to produce fully developed individual categories (Strauss & Corbin, 1990). Table 1 shows the complete list of the categories and subcategories concerned with the foci of interest along with numeric data for each of them. Note that the total frequency is 43 because the study of Dizon (2016) and Suthiwartnarueput and Wasanasomsithi (2012) were tagged under productive skills and knowledge of language categories.

4.1 Demographics of studies

Table 2 shows the distribution of the studies per year with their corresponding demographics. The data reflect a gradual increase in the number of research published from 2010 to 2017. From 1 (2.4%) in 2010, the published studies rose to 7 (17.1%) in 2017; 39 (95.1%) of all the studies are journal articles and 2 (4.9%) are conference papers. In terms of geographical distribution, majority (N = 26) were carried out in Asia, whereas others were carried out in Europe (N = 7), North America (N = 6), Africa (N = 1), and Australia (N = 1). As for the context, most of the studies were conducted in higher education contexts (N = 36) and very few were conducted in secondary (N = 4) and primary (N = 1) education contexts (see Appendix A for the details of each study).

With reference to the research design, 17 of the selected studies were qualitative and 15 used the mixed method design. The least used among the three is the quantitative approach (N = 9). Among the mixed method studies, majority employed descriptive method (N = 9), followed by experimental (N = 5) and correlational approaches (N = 1). The same is true for quantitative studies in which almost half were descriptive (N = 4); the other five were correlational (N = 3) and experimental (N = 2). Among the seven experimental studies in both the mixed method and quantitative designs, only one used a control group. The rest employed a one‐group pretest–posttest design (i.e., preexperimental design). A greater variation was observed in the qualitative studies. Similar to the first two designs, more than half of the studies used a general descriptive approach (N = 10) whereas others used a more specific approach, such as the case study (N = 4), ethnography (N = 2), and grounded theory (N = 1).

In terms of Facebook use, 24 (58.5%) were coded as studies on the formal use of Facebook in formal learning setting. The remaining 17 (41.5%) studies were coded as informal use of Facebook in formal learning setting. None of the studies used Facebook in informal learning settings. These results are similar to the earlier findings of Manca and Ranieri (2016a, 2016b) that studies on Facebook as a learning environment generally used it in formal learning settings.

From the 27 source titles, the most popular publication venues for the reviewed studies are Computers and Composition (4), Computer Assisted Language Learning (3), and Teaching English with Technology (3). Of these, only Computer Assisted Language Learning is classified as a top‐tier journal (Quartile 1) based on 2016 Scimago Journal Ranking. Nonetheless, there were other papers published in top‐tier journals in education and linguistics such as Language Learning & Technology (2), Computers in Human Behavior (1), Canadian Modern Language Review (1), The Internet and Higher Education (1), Language, Culture and Curriculum (1), and ReCALL (1). Findings further reveal that many of the studies were published in regional journals, such as Turkish Online Journal of Educational Technology, Turkish Online Journal of Educational Technology, The Journal of AsiaTEFL, and The Asia‐Pacific Education Researcher.

4.2 Foci of interest

In terms of aims, almost half of the studies zeroed in on exploring the use of Facebook as a supportive and interactive tool for language pedagogy (N = 18). The rest investigated the students' perception of Facebook as a learning environment (N = 13) and its efficacy for language teaching and learning (N = 10). These findings are strongly linked to the distribution of studies according to research design used of the selected studies. It is worthy to note that studies on students' perception and Facebook as a supportive tool typically require descriptive design. Because most of the studies are descriptive in nature, it is expected that majority of the analysed studies would fall under these two aims. Alternatively, studies that test the efficacy of Facebook for language learning usually require an experimental design. Hence, the number of experimental studies and those that test the efficacy of Facebook are almost the same.

As for the focus on specific areas of interest in the field of language pedagogy, the studies were clustered into general language learning, productive skills (i.e., writing and speaking), interaction analysis, knowledge of language (i.e., grammar and vocabulary), intercultural communication and critical pedagogy, metacognition, and Facebook‐based application.

4.3 Issues and challenges in using Facebook as a learning environment

Although the reviewed studies have provided evidence on the positive impact of Facebook on learners and language learning, there remains some reported challenges in its implementation as a learning environment. For instance, students experienced anxiety in posting their work on Facebook (Shih, 2011; Sipacio, 2015; Wichadee, 2013). This anxiety emanated from their lack of self‐esteem, skills, and familiarity with the new rhetorical space.

Students were also confronted with technical difficulties such as formatting, reading texts, limited space for writing comments, and Internet connection (Aydin, 2014; Barrot, 2016; Shih, 2011; Tananuraksakul, 2014) and difficulties in providing feedback and undertaking revision (Aydin, 2014). In some cases, students lamented that Facebook‐based e‐portfolio was logistically demanding such as transforming essays into readable PPTs (Aydin, 2014; Barrot, 2016).

Some studies also reported that using Facebook for academic purpose distracted students from doing their assignments (Shih, 2011) and that not all students were participative in online discussion (Arzu, 2014). Instead, these students merely observed than actually posting content on Facebook. Other students even commented that Facebook was boring when used for academic purpose (Aydin, 2014).

As a rhetorical space, some students considered Facebook as not suitable for writing and learning in general (Kabilan et al., 2010). From the teachers' end, Sipacio (2015) reported that teachers were reluctant to using Facebook on the belief that it is not made for academic purposes and that they were not yet ready to change their traditional practices in language teaching.