Access the full text.
Sign up today, get DeepDyve free for 14 days.
1IntroductionIn recent years, the unparalleled momentum of technological development has pushed both practitioners and researchers to gain a better understanding of the application of technologies in education. Among emerging technologies, learning environments enhanced by virtual reality (VR) have been regarded as positive learning spaces that may further affect learner psychology and actual learning. Mainly empowered by computer-generated simulation technology, VR-enhanced digital environments resemble real or hypothetical worlds in terms of sight, sound, and touch (Alhalabi & Lytras, 2019). They employ displays, tracking, computer-mediated communication (CMC) and other technologies to construct immersive, interactive, and imaginary experiences for users (LaViola et al., 2017). Several review articles have addressed the affordance, opportunities and challenges related to the application of VR in education (Peixoto et al., 2021; Uruthiralingam & Rea, 2020), and novel immersive experiences provided by VR tools have been shown to improve participants’ interest, motivate students to explore new ideas, and thus engage them in active learning (Lau & Lee, 2015; Xie et al., 2019).The potential benefits of VR in the field of language education have been presented by several scholars (e.g. Dalton & Devitt, 2016; Lan, 2015; Wang et al., 2017). Furthermore, previous studies have highlighted “the complex way language learning technologies and virtual environments mediate learners’ acquisition of a second language” (Kramsch & Steffensen, 2008, p. 24), and scholars have emphasised the significance of investigating the process of “second language learning in the specific complexity of innovative computer-mediated contexts” (Mroz, 2015, p. 529). Parmaxi (2020) reviewed 26 studies on the application of VR in language education from 2015 to 2018 and summarised the types of VR technology used in various research settings for learning different target languages. By analysing the potential benefits of VR in language learning, she identified 12 future research directions for VR-supported language education. Drawing upon a review that analysed 21 studies, Li et al. (2021) summarised the potential benefits of VR in improving students’ listening, speaking, writing and cross-cultural communicative competence. Pinto et al. (2021) focused on the gamification of VR and studied 97 publications using a bibliometric approach. Their review study covered the target language, educational settings, and impact of VR games on language education. Peixoto et al. (2021) discussed the features and impact of immersive virtual reality (iVR) in the context of foreign language education. Huang et al. (2021) selected 88 empirical studies and explained the primary uses of augmented reality (AR) and VR technologies, how VR has been used in language teaching and the reasons for improving language learning with VR.As summarised above, several review studies have explored the application of VR in language education. However, only a few coding schemes have been proposed for systematically probing the status and future trends of VR-supported language learning. Therefore, the current review study intends to postulate an analytical framework for understanding the general publication trends, research methods, research foci and benefits and challenges associated with VR-supported language education. This analytical framework could be utilised in future review studies to map the status quo and developmental trends of other emerging technologies in language education practice and research. Since VR has attracted increasing attention among researchers and practitioners in China, this research reviewed 69 empirical studies published between 2010 and 2020 (58 published in international journals and 11 in Chinese journals). By including empirical investigations published by renowned Chinese and international scholarly journals, this study may facilitate a better understanding of the current progress of VR applications in computer-assisted language learning (CALL), both in China and other countries.2VR-supported learning environmentsA VR-supported learning environment generally means “an immersive, three-dimensional (3D), multimedia, and multi-person simulation environment, where each participant adopts an alter ego and interacts with the world in real time” (Wagner & Ip, 2009, p. 250). According to Thorne et al. (2009), the related applications of VR include open social virtualities (e.g. Second Life), commercialised 3D games (e.g. World of Warcraft) and online VR-supported educational spaces (e.g. Quest Atlantis and Croquelandia). While some of the above-mentioned VR environments are experienced through head-mounted displays (HMDs), there are environments that do not require such devices.A new variation of VR is AR, a technology that overlays digital information onto the user’s actual environment. AR has already been used by language educators in a variety of ways (Cheng & Tsai, 2014; Goodwin-Jones, 2016). The latest generation of VR-supported learning environments is empowered by spherical video-based virtual reality (SVVR). It uses a video camera with 360° views and allows designers to integrate spherical videos with VR. In this way, language educators or learners can look around in 360° directions and enjoy authentic sensory involvement (Chien et al., 2020; Walshe & Driver, 2019). Compared with conventional 3D graphics-based VR, which involves costly computer simulations, SVVR is more affordable and presents real-life surroundings to users.Previous research has focused on a set of issues or dimensions for addressing the possibilities, affordance and challenges associated with applying VR to language education. This study conducted a review using the systematic review approach under the PRISMA framework (Moher et al., 2009) and attempted to map the status quo and future trends in applying VR to language education. To fulfil the research purpose, the following four research questions (RQs) were addressed:RQ1:What is the general research trends (including yearly publications, target languages, research settings and types of VR platforms and devices) of the reviewed empirical research?RQ2:What are the research methodologies (including research approaches, research designs, types of data sources and methods of data analysis) employed in the reviewed publications?RQ3:What are the research foci of the reviewed publications?RQ4:What are the potential benefits and challenges of VR-supported language education?3MethodologyTo conduct this review study, we formed a research panel of six researchers: two senior researchers with over 17 years of experience in education and educational research (the first and last author) and four trained assistant researchers (the co-authors) who are pursuing a master’s program in CALL. Based on our research purpose, we divided the review process into three steps (as depicted in Figure 1): journal identification, literature search and data coding and analysis.Figure 1:The process of systematic literature review used in this study.3.1Journal identificationLanguage education supported by VR environments is still an emerging research topic within limited and focused research groups. After a preliminary literature search, we mainly selected research articles published in the field of CALL by 12 international and Chinese journals. Our rationale for selecting the 12 journals is as follows. First, according to previous review studies (e.g. Hsu et al., 2013; Stockwell, 2007), we identified five high-impact scholarly journals that are well-accepted in the field of CALL, namely Language Learning & Technology (LLT), The Journal of EUROCALL (ReCALL), Computer Assisted Language Learning (CALL), System and CALICO Journal (CALICO). The first four journals are included in the Social Science Citation Index (SSCI). Founded in 1983, the CALICO Journal is the first academic journal that specialises in CALL and is considered reputable for its academic devotion and contribution to the CALL community. We included Computers & Education (C&E) as the sixth international journal, since it is the top-ranked journal in terms of impact factor in the field of educational technology and has published several papers on CALL in the last decade. These journals are fully refereed international journals with worldwide authorship and readership. Finally, six journals included in the Chinese Social Science Citation Index (CSSCI) with at least one publication on the application of VR in language education were also selected to better present the current research and development concerning VR applications in China’s CALL community. These six Chinese journals were Modern Educational Technology (MET), China Educational Technology (CET), Technology Enhanced Foreign Language Education (TEFLE), Open Education Research (OER), Distance Education in China (DEC) and e-Education Research (e-ER).3.2Literature searchAfter the 12 journals were identified, we manually screened the studies published by these journals between 2010 and 2020 based on the titles, abstracts and keywords issued by the official websites of the 12 journals. Following our previous studies (Zheng et al., 2019a, 2019b), we used two sets of keywords for literature screening: (1) 3D virtual/interactive learning environments, VR, virtual worlds, virtual characters, avatars and Second Life; (2) foreign language (e.g. English, French), language teaching (teachers), language learning (learners), language education and computer-assisted language learning. Articles containing information related to these sets of keywords were identified for further analysis. Moreover, we only retained original research articles and excluded book reviews, commentaries and editorial materials. A total of 69 empirical studies were eventually retained based on their relevance to our research purpose (see a brief summary of the 69 studies in Appendix I). The electronic files of the 69 full-length articles were then downloaded for further analysis.3.3Data coding and data analysisThe process used in this study for coding and analysis is depicted in Figure 2. We divided the research panel into two independent groups (each with two assistant researchers and one senior researcher). As shown in Figure 2, data coding and analysis included two stages: Stage 1 as the preliminary coding and analytical stage and Stage 2 as the formal coding and analytical stage.Figure 2:The process of data coding and data analysis used in this study.3.3.1Preliminary coding and analytical stageAt the preliminary coding and analytical stage, the research panel referred to the coding scheme and analytical framework proposed by previous studies (Chai et al., 2013; Macaro et al., 2012), as well as the categorisation of research topics on VR-supported language learning in the literature (Lin et al., 2013; Zheng et al., 2019b). We then postulated a tentative coding framework to address the aforementioned four research questions. The individual researchers in the two groups were asked to code the same 12 articles (one article from each of the 12 journals) independently. After completing the coding, the tentative coding framework was revised based on a panel discussion. Disagreements were resolved, and the tentative coding framework was revised after a panel discussion guided by senior researchers. To further guarantee the validity and reliability of the analytical framework, two CALL experts were invited to inspect the categories and sub-categories. At the end of Stage 1, a revised coding framework with categories, sub-categories and definitions was then postulated for Stage 2, i.e. formal literature coding and analysis.3.3.2Formal coding and analytical stageAt the formal coding and analytical stage, the researchers in the two groups followed the revised coding framework and performed a content analysis for the 69 publications independently. After completing the coding, the research panel compared the resulting data for RQ1 and RQ2, which are mainly statistical data. The average inter-rater reliability among the two groups was above 0.95, which indicates good agreement.The resulting themes for RQ3 and RQ4 were derived from the articles holistically, with keywords, phrases and sentences identified to reflect the salient content that researchers have been discussing. The panel compared the qualitative data and came to an agreement after several rounds of reading, re-reading and discussion led by the senior researchers. After the panel agreed on all the coded categories, sub-categories and specific items, the finalised coding framework was then formed with four main categories and 19 sub-categories (as shown in Table 1).Table 1:The coding framework for reviewing literature on VR-supported language learning.Main categoriesSub-categoriesResearch question 1 (RQ1)General research trends1. Year of publication2. Authors’ geographic distribution3. Research sites4. Research settings5. Target languagesResearch question 2 (RQ2)Research methodologies6. Research approaches7. Research designs8. Data collection9. Data analysisResearch question 3 (RQ3)Main research foci10. Affordance of VR for linguistic knowledge and skills11. Affordance of VR for non-linguistic knowledge and skills12. Language learner characteristics in VR-supported learning environments13. Language teachers in VR-supported learning environments14. The design of language learning tasks with virtual reality15. The design and development of VR-supported learning environments16. The nature of interactions in VR-supported learning environmentsResearch question 4 (RQ4)Potential benefits and challenges17. Features of effective VR-supported learning environments18. Affordance of VR for language learning19. Challenges of VR for language learningDuring the data coding and analysis phase, we used the Statistical Package for the Social Sciences (SPSS 25.0) to process the statistical data and analyse descriptive data (such as the amount of publications and its percentage in the total number of publications) for RQ1 and RQ2. For RQ3 and RQ4, we employed NVivo 11 to annotate the keywords, phrases, and sentences, and to make further categorisations.4Results and discussion4.1General publication trends4.1.1Number of empirical studies published from 2010 to 2020As indicated by Figure 3, 69 empirical studies were published by the 12 journals between 2010 and 2020, among which 58 articles were published by six international journals and 11 articles by six Chinese journals. In the last 11 years, Computer Assisted Language Learning, Computers & Education and ReCALL published the largest number of studies. In Figure 3, the yearly number of publications in different journals is shown in solid lines, whereas the total number of publications is indicated by the dotted line. The number of articles published on the integration of VR into language learning is still small owing to the complexity of using VR in language education (e.g. Kramsch & Steffensen, 2008; Mroz, 2015). The largest number of empirical studies were published in 2019 and 2020, which indicates a boost in applying VR technologies to language learning in recent years. With an accumulated understanding of designing language lessons in VR environments, coupled with the advancement of the technology itself, we are certain that more empirical investigations will be conducted in the near future.Figure 3:Publication numbers for the 12 journals (2010–2020).4.1.2Research sites and settings of the studies between 2010 and 2020Based on the authors’ geographical distribution and the research sites for conducting the research, we found that scholars from Chinese Taiwan have contributed a large portion of the empirical studies over the past 11 years (20 in international journals). Moreover, scholars from the United States (9 studies) and Australia (5 studies) are also active in the research on VR-supported language learning. As shown in Figure 4, a total of 40 empirical studies (35 in international journals and 5 in CSSCI journals) were conducted in higher education settings, indicating the relatively popular use of VR technologies among adult language learners. 11 studies were conducted in elementary (10 in international journals and one in CSSCI journals) and another 11 in secondary education settings (seven in international journals and four in CSSCI journals). As illustrated by Figure 4, studies published in international journals have also paid attention to the setting of social adult education. The presence of a smaller number of studies conducted in K-12 settings is consistent with previous review studies (e.g. Lan, 2015). Although several studies have focused on the possibility of applying VR technologies, particularly game-based language learning, in primary or secondary educational settings, only a few studies have focused on improving young learners’ self-directed learning in VR settings (e.g. Lan, 2015). For the design of VR-supported situational learning environments for primary school students, Lan (2015) summarised three principles, namely, “individuality”, “adaptability” and “scaffolding”, that provide suggestions for further improving the application of VR in K-12 education.Figure 4:The research settings of the studies.4.1.3Target languages studied in the reviewed publications between 2010 and 2020As exhibited in Figure 5, English is the dominant target language under research, featuring 35 international journals and 8 CSSCI journals. The popularity of English in the academic, economic and technological domains is well established (e.g. Lan, 2015; Lan et al., 2007; Su, 2006). A central topic in CALL is helping learners master English by leveraging emerging technologies. It is noteworthy that 10 studies (7 published in international journals and 3 in CSSCI journals) explored learning Chinese in VR-supported learning environments. Chinese language and culture are becoming increasingly popular, and we expect that the number of related studies will increase in the coming years. Five studies reported the use of VR-supported language education for speakers of various languages belonging to diverse cultural backgrounds. Levak and Son (2016) investigated the effectiveness of using Second Life and Skype in developing listening comprehension among learners of Croatian or English. Canto et al. (2013) described a two-year cross-cultural language exchange project in which 60 interactional tasks were implemented by learners who had Dutch, Portuguese, Russian, or Spanish as their native languages. In their study, rich interactions for exchanging social and cultural meanings were found between language learners and native-speaking in-service teachers.Figure 5:The target language investigated by the empirical studies (2010–2020).4.1.4Types of VR platforms and devicesAmong the 69 journal articles, 66 (55 published in international journals and 11 in CSSCI journals) specifically indicated that they employed VR technologies. Most studies (26 published in international journals and 2 in CSSCI journals) explored learners’ language learning in Second Life, one of the most popular commercialised VR platforms. As indicated by Stevens (2006), Second Life encourages users’ creativity and has the potential to bring about changes to education. Moreover, the emergence of several self-designed VR-supported learning environments adds to the diversity of users’ experiences in VR-supported learning environments. Ten studies (8 published in international journals and 2 in CSSCI journals) used AR to support learners’ language learning (e.g. Chen et al., 2019; Cheng & Tsai, 2014; Danaei et al., 2020). Several studies conducted in other countries employed open-source VR platforms, such as Open Simulators (e.g. Berns et al., 2013; Dalton & Devitt, 2016) and Open Wonderland (Chen et al., 2011). Figure 6 indicates that two empirical studies have applied SVVR for several personalised educational purposes. Immersive learning environments created using SVVR can support the cultivation of students’ enthusiasm for participation and writing literacy. SVVR has broad application prospects in the field of writing learning (Chen et al., 2020). Since most commercialised immersive environments were not primarily developed for educational purposes (e.g. Berns et al., 2013), we propose that language educators need to work closely with educational technologists to create more learner-oriented VR environments that would facilitate learners’ second language acquisition (SLA).Figure 6:Types of VR-supported learning environments in the empirical studies (2010–2020).4.2Research methodologies4.2.1Research approachesBased on Creswell’s (2010) framework, we identified 3 research approaches and 6 main types of research designs among the 69 studies. As indicated in Table 2, 14 studies conducted qualitative inquiries, among which 13 articles used case studies and one employed grounded theory. Another 22 studies adopted a quantitative research approach, including two corpus analyses and 20 experimental designs. Most studies adopted the mixed-methods approach: 33 of the 69 studies adopted mixed methods for data collection and data analysis, among which 15 studies used convergent mixed-methods designs and 9 studies implemented explanatory sequential mixed-methods designs.Table 2:Research approaches and research designs of the reviewed studies (2010–2020).Research approachesResearch designsNumber of studiesQualitative researchCase studies13Grounded theory1Quantitative researchCorpus analysis2Experimental research20Mixed researchAction research3Convergent mixed methods design15Explanatory sequential mixed methods design9Exploratory sequential mixed methods design2Embedded or other unspecified mixed methods design4A possible reason for the wide use of mixed methods in these studies is the complex nature of learner interactions in VR-supported learning settings. Researchers have indicated the complicated and even chaotic process of learners’ SLA within virtual learning environments (e.g. Mroz, 2015). Mixed methods may allow researchers to better discover a holistic and complicated process of L2 learning with VR. The triangulation of mixed types of data may further verify the effects of VR on language-learning outcomes.4.2.2Data collection methodsAs indicated in Figure 7, the reviewed studies employed a variety of data sources. For instance, Wigham and Chanier (2013a) used six types of data sources for analysing learners’ verbal and non-verbal communication in Second Life, including computer messages (screen recording), audio-taped materials, surveys, observations, online interviews and corpora of learners’ communication acts. Another typical study was conducted by Ho et al. (2011). They included mixed types of data sources (five types of data in total) in their investigation and analysed interviews, notes of classroom observations, student-generated artefacts, students’ reflections and survey responses. Through a combined analysis of mixed types of data, they described and further evaluated the effectiveness of implementing a self-designed virtual museum in a language curriculum.Figure 7:Frequency of data collection approaches used in the empirical studies (2010–2020).The most frequently used data collection method was testing, which was used by 26 studies published in international journals and eight in CSSCI journals. Moreover, 26 studies reported on one-to-one interview data (25 studies conducted person-to-person interviews, whereas one study conducted interviews by email). Two studies conducted focus group interviews. Thirty studies used surveys or questionnaires for data collection, 28 used comparative groups in their research design and 18 adopted observations for collecting data when learners were engaged in VR environments. Eight studies used different types of computer messages, such as data sources. 8 studies included video-taped sources, and 6 used audio-taped sources for data analysis. Fourteen studies used private documents, such as learning journals or reflective reports, as data sources. Figure 7 reveals that several studies conducted in China and other countries adopted testing, surveys or questionnaires and interviews as the primary data collection method. The data collection methods employed by studies published in CSSCI journals were less diverse than those adopted by studies published in international journals.4.2.3Data analysis methodsThe studies reviewed herein employed various quantitative and qualitative data analysis approaches (refer to Figure 8). Of the 36 research articles that employed qualitative data analysis (34 published in international journals and two in CSSCI journals), 23 studies (21 published in international journals and two in CSSCI journals) used corpus-based content analysis, including interaction analysis and discourse analysis based on chat scripts or recorded interactions, thereby exploring the characteristics of interactions produced in the virtual world. Nine studies derived descriptive summaries by analysing observation data, by identifying the themes that emerged from interviews (e.g. Rama et al., 2012) or by analysing learner behaviours or problems that occurred during the application of VR in language education (e.g. Lan, 2015). In addition, three studies performed thematic analysis (e.g. Ebadi & Ebadijalal, 2020), and one study conducted interpretive analysis (Xie et al., 2019). Concerning quantitative data analysis (refer to Figure 9), 36 studies (30 published in international journals and six in CSSCI journals) used descriptive quantitative analysis, such as the frequency, mean, deviation or comparison of means based on survey responses. Fifteen studies employed the analysis of variance (ANOVA), and five studies employed the analysis of covariance (ANCOVA) to probe the possible effect of VR environments on students’ learning achievements (e.g. Cheng & Tsai, 2014; Hsu, 2017; Lan, 2016). Twenty-two studies conducted t tests (e.g. Chen et al., 2011; Franciosi et al., 2016; Yang et al., 2020), three used the Chi-square test (e.g. Franciosi et al., 2016; Lan, 2014), 4 used the Wilcoxon signed-rank test (e.g. Chen et al., 2019; Shih, 2014) and three conducted reliability tests (e.g. Chen et al., 2011; Collentine, 2011). As shown in Figures 8 and 9, various data analysis methods were used to enhance the validity and reliability of the research. It can be concluded that the studies published in international journals employed richer types of data analysis compared with those published in CSSCI journals. Descriptive statistics, t-test and ANOVA were the most frequently adopted data analysis methods across the 12 journals. Future research may consider how to triangulate mixed types of data for a more refined understanding of how VR may affect or improve language education.Figure 8:Frequency of qualitative data analysis approaches used in the empirical studies (2010–2020).Figure 9:Frequency of quantitative data analysis approaches used in the empirical studies (2010–2020).4.3Research foci in the reviewed studiesAs illustrated in Table 3, our analysis extracted a total of seven research foci related to the investigations of VR-supported language education, including the affordance of VR for improving linguistic knowledge and skills (42 studies), the affordance of VR for improving non-linguistic knowledge and skills (25 studies), language learner characteristics in VR-supported learning environments (40 studies), language teachers in VR environments (3 studies), the design of language learning tasks with VR (12 studies), the design and development of VR environments (8 studies) and the nature of interactions in VR environments (12 studies).Although most studies fall into one of the seven categories of the extracted research foci mentioned above, some studies have more than one research focus. For instance, Jauregi et al. (2011) reported the potential benefits of VR-supported projects for solving learners’ cultural, linguistic, interpersonal, and motivational issues, thereby covering several research foci, as summarised above. Tseng et al. (2020) investigated the effects of VR on young learners’ vocabulary acquisition and specifically explored the interplay between learner autonomy and collaboration to examine the effects of the socio-cultural dimension of such practice. Ebadi and Ebadijalal (2020) found that VR has a positive effect on participants’ knowledge, self-efficacy, and cultural sensitivity.Table 3:Research foci of the 70 empirical studies (2010–2020).CategoriesSub-categoriesNumber of studies1. Affordance of VR environments for linguistic knowledge and skills (42 studies)Listening2Speaking9Writing6Translation/interpreting1Vocabulary retention8Overall language proficiency162. Affordance of VR environments for non-linguistic knowledge and skills (25 studies)Critical thinking2Cognitive attainments4Willingness to communicate (WTC)2Communicative skills/strategies12Collaborative learning skills63. Language learners’ psychological characteristics in VR environments (40 studies)Learner perceptions/attitudes17Learner autonomy3Learner motivation13Learner engagement6Learner anxiety7Learning styles1Self-efficacy5Self-regulation1Boredom1Flow state24. Language teachers in VR environments35. The design of language learning tasks in VR environments126. The design and development of VR environments87. The nature of interactions in VR environments124.3.1Research foci one: affordance for linguistic knowledge and skillsThe ecological notion of “affordance” means something available to organisms for them to use and accomplish a certain goal (Gibson, 1979, p. 128). Drawing upon the ecological perspective, Van Lier (2000) proposed changing “input” to “affordance” in SLA. Similar to the mechanism wherein a leaf provides different affordances for different organisms, technologies offer a variety of affordances to learners and teachers in language education (Liu & Chao, 2018, p. 70). Among the 69 studies, 42 focused on the affordance of VR in enhancing language learners’ linguistic knowledge and skills (34 published in international journals and 8 in CSSCI journals). With VR technologies, learners have opportunities to encounter situational language exchanges through participation and navigation (Tai et al., 2020). VR is also considered to be one of the easiest and most effective learning environments for players to communicate and have fun (York et al., 2020). As indicated in Figure 10, 16 articles explored the potential benefits of VR in improving language learners’ overall language proficiency, language performance or target language output. Other articles addressed improving particular types of linguistic skills through accomplishing learning tasks in VR environments, such as listening (2 studies), speaking (9 studies), writing (6 studies), interpreting (1 study) and vocabulary retention (8 studies). VR-mediated vocabulary learning is enjoyable and motivating since it provides a learning setting with comprehensible input, simultaneous interaction and instant feedback (Tai et al., 2020). Therefore, students can access meaningful content topics to improve their foreign language vocabulary acquisition (Alfadil, 2020).Figure 10:Empirical studies on the affordance of VR for linguistic knowledge and skills. Note: “The ‘C’ in the brackets indicates studies indexed by CSSCI, and ‘I’ in the brackets indicates studies published in international journals.”4.3.2Research foci two: affordance for non-linguistic knowledge and skillsAs indicated by Figure 11, 25 of the 69 studies explored the affordance of VR for enhancing learners’ non-linguistic knowledge and skills (23 published in international journals and 2 in CSSCI journals). Among the 25 studies, 12 articles addressed fostering language learners’ communicative competence in VR environments. Communicative competence is a critical objective in foreign language teaching (Tang et al., 2016). It not only indicates learners’ knowledge about the language but also assesses their competence in using the language appropriately in real-life situations. In addition, 6 studies discuss learners’ collaborative learning in VR-supported environments. VR environments provide students and teachers with diverse scenarios for communication, through which they can participate in meaningful and collaborative learning (Doumanis et al., 2019). Other benefits of VR environments include promoting learners’ critical thinking (Mroz, 2015), their cognitive attainment (Cheng & Tsai, 2014; Hsu, 2017; Li & Wang, 2016) and their WTC. As suggested by previous studies, WTC is a positive state of mental preparation when language learners get the chance to communicate (Kruk, 2019). Students with higher WTC have been observed to be more successful in the process of language acquisition. Our reviewed articles show that the application of VR further enhances learners’ colloquial skills and WTC in a light-hearted and reassuring environment (Ebadi & Ebadijalal, 2020; Kruk, 2019).Figure 11:Empirical studies on the affordance of VR for non-linguistic skills. Note: “The ‘C’ in the brackets indicates studies indexed by CSSCI, and the ‘I’ in the brackets indicates studies published in international journals.”To summarize, a very small number of empirical studies published by CSSCI journals explored language learners’ non-linguistic knowledge and skills, and there is still a need to improve our understanding of Chinese learners’ communicative strategies or collaborative learning in VR-supported environments. Generally, only a few investigations have probed into learners’ critical thinking or willingness to communicate (WTC) with the support of VR, and we are looking forward to more in-depth explorations in the future.4.3.3Research foci three: language learners’ psychological characteristicsSeveral studies reviewed herein explored how VR tools improve language learners’ learning outcomes by influencing their psychological characteristics in VR-supported learning environments. As shown in Figure 12, the most investigated topics are language learners’ perceptions of or attitudes towards their learning experience in VR environments (17 studies) and language learners’ motivation or interest (13 studies). Seven studies addressed language learners’ anxiety in VR-supported learning environments (e.g. Hsu, 2017; Wehner et al., 2011), and 6 studies explored learner engagement (e.g. Chen et al., 2020; Chen & Kent, 2020; Liang, 2012). Four studies explored learner autonomy (Collentine, 2011; Tseng et al., 2020; Zhu & Zhang, 2011) and self-regulation (Chen & Hsu, 2020). As observed in previous studies conducted on CALL, learner psychology has always been a highly investigated area, and within this topic, learner attitude, motivation, anxiety, engagement and autonomy constitute a large proportion of the investigations. In general, only a few studies were published by CSSCI journals on learners’ psychological characteristics in VR settings. Since individual differences are crucial for further understanding the psychology of language learners, we expect more investigations focusing on this aspect in the future.Figure 12:Empirical studies on language learners’ psychological characteristics. Note: “The ‘C’ in the brackets indicates studies indexed by CSSCI, and the ‘I’ in the brackets indicates studies published in international journals.”4.3.4Research foci four: role of teachers in VR environmentsThree studies investigated teachers’ roles and teacher training in VR-supported language education. Ernest et al. (2013) reported a small-scale teacher development project in which 20 language instructors were trained to master the skills necessary for promoting learners’ online collaborative learning in virtual environments. Wang (2015) employed discourse analyses and the teachers’ speech act in Second Life and revealed teachers’ critical roles in designing collaborative tasks, providing input, giving instructions, managing student collaboration and providing feedback for facilitating students’ language learning in different task phases. Kozlova and Priven (2015) explored the effectiveness of a collaborative situated learning approach for training language educators to teach in 3D virtual worlds. They found that VR-supported teacher training allowed participant teachers to improve their ability to construct personalised models of tasks and improve teaching practices, which further helped the teachers achieve their pedagogical goals. Wang (2017) invited teachers of the Chinese language to co-design AR-supported tasks for improving the writing skills of high-school students. The findings revealed the teachers’ role in promoting Chinese language learners’ writing performance in terms of controlling the content, organising the article and choosing proper words when they were involved in writing. Owing to the complexity of integrating VR technologies into the pedagogical practices of language education, more empirical studies are still needed to further elaborate teachers’ roles and how trainee teachers can be better supported to achieve their pedagogical objectives. Wang (2015) classified teachers’ roles into traditional roles (e.g. monitor, motivator and language guide), technical roles and social roles of task-based language teaching (TBLT) in virtual environments. In particular, the technical role and social role, which require teachers to deal with technical glitches and build appropriate social cohesion, are critical in VR-based learning environments. Despite having command over the technology, to apply VR, language educators also need to be equipped with skills such as framing and designing collaborative tasks, providing authentic oral input and giving feedback to students for implementing effective task-based instruction in 3D virtual worlds (Kozlova & Priven, 2015). In addition, Wang (2017) suggested that teachers may participate in designing AR-enhanced educational materials to inspire students’ learning in diverse settings.4.3.5Research foci five: task types and designsTBLT has been widely acknowledged as an effective means of designing conventional classroom teaching (e.g. Ellis, 1994, 2003; Nunan, 2004; Pica et al., 1993). Twelve of the 69 articles reviewed herein reported task types and designs in VR environments. For instance, Collentine (2011) designed two language learning tasks by integrating 3D exploration and synchronous computer-mediated communication, and they further investigated how the completion of these two tasks may affect Spanish language learners’ linguistic complexity and language accuracy. As displayed in Figure 13, seven studies that harnessed Second Life and designed tasks for language learners reported positive effects of proper task designs in a 3D virtual world (e.g. Jauregi et al., 2011; Lan et al., 2016). Van Ginkel et al. (2019) asked learners to practice their presentations in a virtual environment and noted that receiving immediate system feedback may further improve students’ cognition, behaviour and presence. Recently, Chen (2020) found that pre-task planning should be embedded in task-based instruction to enhance learners’ performance. He recommended that tasks should be oriented to the actual world and should aim to improve learners’ world knowledge and cultural repertoires, which could positively impact their virtual learning.Figure 13:Empirical studies on the task types and designs in VR learning environments.Note: “The ‘C’ in the brackets indicates studies indexed by CSSCI, and the ‘I’ in the brackets indicates studies published in international journals.”4.3.6Research foci six: design and development of learning environmentsEight studies were identified to explore the design and development of VR-supported learning environments (Figure 14). Among them, four studies implemented self-designed VR-supported learning environments (e.g. Berns et al., 2013; Li & Wang, 2016), and four studies employed existing VR platforms such as Second Life (e.g. Guo et al., 2012; Lan, 2015). Yang et al. (2010) designed an English learning system (PILE system) and demonstrated through comparative group research that the system was beneficial for maintaining students’ long-term learning and improving their learning motivation. Ho et al. (2011) reported a student-constructed virtual museum of language curriculum (MUSE) and verified the variability of the system in enhancing collaborative learning skills and language learning motivations through the implementation and evaluation of the system. Berns et al. (2013) designed a game-like virtual platform (VirUAM) with several virtual islands built inside and their aim was to provide basic training for different levels of German language learning. For self-developed VR-supported learning environments, Liang (2012) revealed that students’ collaborative interaction activities could help create and develop language-associated tasks within online role-playing games, as well as carefully designed online role-playing games augmented by specific contexts, which would facilitate learners’ language learning. Recently, Lan (2015) evaluated the effect of virtual immersive contexts on primary students’ English learning through self-developed virtual contexts in Second Life and found that the developed virtual contexts could serve as diversified language learning opportunities compared with regular classroom schedules. Based on the theory of embodied cognition, Li and Wang (2016) designed an online virtual English learning community in a primary school. It changed the traditional online learning environment, highlighted the embodiment of learning methods, provided diversified online learning methods and constructed a new situational online learning method, which facilitated better language learning.Figure 14:Empirical studies on the design and development of VR learning environments. Note: “The ‘C’ in the brackets indicates studies indexed by CSSCI, and the ‘I’ in the brackets indicates studies published in international journals.”4.3.7Research foci seven: nature or features of interactionsTwelve of the 69 studies reported findings specifically related to the nature or characteristics of interactions when learners were engaged in a learning environment enhanced by VR (Figure 15). In sociocultural theory, interaction is considered ‘an essential force rather than as merely a helpful condition for learning’ (Saville-Troike & Barto, 2017, p. 118). However, in many English as a foreign language (EFL) contexts, decontextualised language learning may cause a lack of communication or authentic ways of using the target language (Lee & Park, 2019). VR offers learners a positive and stimulating environment for promoting social interactions through collaborative dialogues in their target language (Canto et al., 2013; Jauregi et al., 2011; Peterson, 2012). Morton and Jack (2010) explored learners’ response type by analysing their utterances in a CALL program supported by applications combining virtual agents with automatic speech recognition technology. Yang et al. (2010) discussed physical interactions when learners engaged in VR-technology-enhanced learning situations. Cheng and Tsai (2014) revealed four patterns of child-parent reading behaviours with AR picture books. Melchor-Couto (2018) designed four tandem oral interaction activities to explore the effect of anonymity in virtual worlds on language learners’ oral interactions. Since VR-supported learning environments are very complex and involve both verbal and non-verbal communicative instances, four studies also proposed specific analytical frameworks for understanding the features of interactions in VR environments (e.g. Mroz, 2015; Peterson, 2010; Shih, 2014; Wigham & Chanier, 2013b). These frameworks provide models for further understanding learners’ interactional behaviours in VR environments.Figure 15:Empirical studies on the nature of interactions in VR learning environments.Note: “The ‘C’ in the brackets indicates studies indexed by CSSCI, and the ‘I’ in the brackets indicates studies published in international journals.”4.4Potential benefits and challenges of VR environments for language learning4.4.1Affordance of VR-supported language educationTable 4 briefly summarises the affordance of VR environments for language education. The first and foremost feature of the effectiveness of VR-supported language learning lies in its novel and immersive user interface empowered by 3D simulation technology or 360° spherical video technology. Learners were able to create their own virtual characters (e.g. Peterson, 2010), improve their enjoyment in participation (e.g. Lan, 2014; Yang et al., 2010) and enhance their motivation in learning (e.g. Berns et al., 2013; Yang et al., 2020). VR allows learners to situate themselves in different cultural or geographical contexts. Therefore, it can overcome the limitations of physical classrooms and encourage language learners’ telepresence in semi-realistic or quasi-realistic contexts (e.g. Peterson, 2012; Shih, 2015).Table 4:Features of effective VR-supported learning environments and their affordance.Features of effective VR learning environmentsAffordance for language learners1. 3D user interface with immersive simulationIncreasing level of immersionImproving motivation (intrinsic interest)Enhancing engagement in learning tasksOffering contextualised learning experienceProviding joyful learning experience2. Anonymity afforded by the avatarsIncreasing sense of telepresenceDecreasing learner anxietyEnhancing self-confidence3. Synchronously interactive CMC environmentsIncreasing social interactionFacilitating collaborative learningEnhancing linguistic performanceImproving engagement in authentic culturesOvercoming spatial or temporal limitationsSecond, the avatar, a cyber-self of the participant in VR environments, enables language learners to be anonymous and thereby avoid experiencing embarrassment, which further promotes learner engagement and self-efficacy in language learning (e.g. Chen, 2020; Melchor-Couto, 2018). The avatar provides students with “masked security”, which enables them to use English outside their comfort zones and thereby avoid the social anxiety they may have to face in person in real-life situations (Chen, 2014; Jung, 2002; Yamazaki, 2018). As suggested by prior research (Ebadi & Ebadijalal, 2020), VR enables language learners to practice in a relatively safe and relaxed environment, which may make them feel more comfortable because of the absence of others’ judgement.The third innovative feature of VR-supported learning environments involves synchronously interactive environments constructed using CMC technology. Learners are equipped with multimodal methods of communication that integrate audio, visual and textual modes for both synchronous and asynchronous interactions (Chen, 2016; Wigham & Chanier, 2013a, 2013b). Social interaction is considered the fundamental way of SLA (e.g. Frawley & Lantolf, 1985; Lei, 2008; Yu & Lee, 2016). The multiple modes of communication in the VR space can assist learners in overcoming the spatial or temporal limitations of the physical world and provide learners with more authentic contexts for target language learning (Lan, 2015; Shih, 2015). Authentic target language interactions and immediate feedback from virtual characters can help learners enlarge their vocabulary and pragmatic competence (Tai et al., 2020).4.4.2Challenges of VR-supported language educationIn the articles reviewed herein, scholars reported different challenges concerning the effective use of VR-supported environments in language education, including unstable Internet connections caused by limited bandwidths (e.g. Chen, 2016; Lan, 2014; Levak & Son, 2016) or the inaccessibility of VR platforms due to firewalls (e.g. Peterson, 2010). Since VR-supported language learning environments usually require a specialised operating system and related equipment, such as headsets or 360° cameras (e.g. Chen, 2016; Lan, 2014), they are still not available to all those who are interested in learning with VR. Making these devices more affordable and accessible may become a critical issue in the next few years. Conducting learning tasks in VR environments can at times be a little complicated (e.g. Kozlova & Priven, 2015; Lan, 2014), and language learners need to learn how to navigate avatars in the VR environment. Participants reported frustrating experiences when they encountered glitches when navigating to complete learning tasks. Several learners also reported symptoms of motion sickness in VR, such as nausea and dizziness (Zheng et al., 2021). Thus, making VR-supported learning environments easier to use with more effective navigating operations may become another critical issue for enhancing the affordance of VR environments (e.g. Chen, 2016; Morton & Jack, 2010).The final challenge reported by the reviewed studies focuses on the extra effort that language teachers have to make, such as designing learning tasks, managing in-class activities and interacting with learners during their spare time. For instance, Chen et al. (2019) reported that using Google Earth Virtual Reality requires better classroom supervision and providing more attention to learners for assisting them in their writing, as they may keep looking around, explore everything in the environment and forget to do what they are asked to accomplish. It was equally challenging for teachers to organise a virtual class smoothly through the application of VR, as a significant amount of time was spent familiarising students with the operations of the VR equipment and gathering them all in one place before teleporting them to the virtual environment for completing tasks (Chen & Kent, 2020). These extra efforts may add to teachers’ workload (e.g. Kozlova & Priven, 2015; Lan, 2014; Wang, 2015), and how technology and classroom practitioners could work together more effectively still remains a challenging issue for CALL.5ConclusionThe current review study analysed 69 empirical studies that explored language education supported by VR. We proposed a coding framework for reviewing these studies and addressed four research questions. Similar to other types of technological innovations, VR-supported learning environments are promising for language education, but they also pose a variety of challenges. Currently, the critical issue is not “whether” this innovative technology could work for language education, but “how” and “why” it could improve the effectiveness of language learning through VR. Since VR at this time is not advanced enough to render all the graphical information from the real world into the virtual world, VR-supported environments have not fulfilled their full potential. Although SVVR has made a realistic sensory experience possible, only a few studies have employed it to develop learning materials or to design related language learning tasks thus far.The ultimate purpose of VR-supported environments for language learning is to promote learners’ linguistic and non-linguistic skills, and proper task designs are the key factors for achieving effective learning in VR. As summarised in the reviewed articles, immersive simulation of real-life tasks in VR environments decreases virtual distance and optimises the implementation of a task-based syllabus design (González-Lloret, 2015; Ortega & González-Lloret, 2015). However, pedagogical challenges also exist, mainly depending on how language teachers could effectively use VR environments, design appropriate language learning tasks and successfully integrate them into the current language classrooms. We are still at the dawn of applying VR in language education in the digital age. Researchers, teachers and VR technological specialists need to collaborate to design appropriate learning tasks aimed at improving language learners’ linguistic skills and cross-cultural communication competence.This study only reviewed empirical studies in designated and CALL-related journals published in China and other countries between 2010 and 2020. In our future study, it will be important to include more studies on VR-supported language education published by linguistics and educational technology journals. It is suggested that future studies expand their literature search to other databases, such as Google Scholar, Web of Science and IEEE, which can ensure exhaustive inclusion of relevant literature. It may also offer a broader overview of the status quo of VR-supported language education. Besides, the current review only summarised the general publication trends, research methods, research foci, promises and challenges of applying VR in language education. Future research could investigate the effectiveness of VR-supported language teaching and learning based on meta-analysis and analyse the theoretical foundations for better guidance of instructional design with VR.
Journal of China Computer-Assisted Language Learning – de Gruyter
Published: Dec 1, 2022
Keywords: affordance; language education; systematic review approach; virtual reality
Access the full text.
Sign up today, get DeepDyve free for 14 days.