Exploring the trends of educational virtual reality games: a systematic review of empirical studies

The educational process requires the learner to grasp and comprehend abstract concepts and appreciate scenarios, as well as to understand situations which are far removed from the confines of the classroom. Duffy and Jonassen (1992) argue that the teaching of abstract phenomena, analogy and lived experiences are used to describe and ease through abstract concepts from commonly observable reality. However, over the last decade, the application of emerging technologies in education have revolutionized the pedagogical or teaching processes in the classroom, to explain better and to provide comprehension of abstract concepts.

VR is an emerging technology that has gained traction in education. Educationists have discovered that VR allows the user to interact with a computer-generated 3D model or virtual environment (Christou, 2010) which fosters the understanding of the imaginary world on a realistic scale. This makes it a useful tool for teaching and learning by transforming the way educational content is delivered (Hentsch, 2018). A study by Lee and Wong (2008), described VR as a technology that aids student learning through the visualization of information and engagement, which affords a learner with the opportunity to experience subject matter or concepts that are not easily discernible.

VR does not only relate positively to education but aligns with the constructivism school of thought, which stipulates that humans construct knowledge by learning from experience. The constructivist theory advocates that learning through interaction with sensory data allows knowledge construction from experience for which VR is suited. Christou (2010), chronicled broad application areas of VR in educational contexts used in the enhancement of core curricula in schools and colleges, application of VR in museums, edutainment, demonstrations, simulation, and training.

In the past few decades, game-based learning has been an integral part of the educational process (Boboc, Orzan, Stoica, & Niculescu-Ciocan, 2018; Hwang, Wu, Chen, & Tu, 2016; Xenos, Maratou, Ntokas, Mettouris, & Papadopoulos, 2017). EVRGs are designed to assist users to grasp the concepts of a specific subject, to expand their knowledge, and to facilitate participation. A typical example of educational virtual reality game is GridlockED, which was developed by Tsoy et al. (2019) as a collaborative learning game, targeted for medical trainees, to acquire the skills on how to treat and triage a patient. Modern educational games are played using high-end video technology such as stereoscopic 3D, or through a Head-Mounted VR environment (Snowdon & Oikonomou, 2018). These technologies use a spatial depth on the screen and offer rich learning experiences for the user.

A study by Sampaio, Ferreira, Rosário, and Martins (2010) described 3D technology as that which creates spatial depth with 3D pop-up visualizations so that objects contained in the game may appear closer to the player with the use of dedicated 3D-capable glasses. In comparison to the 3D technology, Triberti, Villani, and Riva (2016) described VR as a computer-simulated reality that uses a 3D environment whereby the player interacts with using a specialized head-mounted display (e.g. Oculus Rift, Samsung Gear VR, Google Cardboard etc.) that shows visual effects to the eyes. Unlike 3D, where the user is a just a viewer, VR allows for the user to become part of a story, offering an untethered immersive experience (Bradshaw, 2016). Therefore, a video game in VR is more realistic than one played in a 3D condition. Hence, VR technology could be applied to complement 3D modelling, ensuring better communication both in an educational or vocational training (Bradshaw, 2016). A comparative study by Roettl and Terlutter (2018) on 237 players showed that the participation and cognitive load was much higher in players using VR than just 3D. Recently, researchers have critically debated on the effects of 3D and VR technologies on issues such as adverse effects like social isolation (Nicas, 2018), discomfort, eye fatigue, and headache when using these technologies for an extended period (Bradshaw, 2016; Roettl & Terlutter, 2018; Sharkawi, Ujang, & Abdul-Rahman, 2008).

Since this article focuses on exploring current EVRGs, we consider it vital to summarize in a tabular form the existing efforts in this direction. Table 1 presents the previous attempts to review works in EVRGs.

As can be observed from the existing studies reviewing EVRGs (Table 1), the field of VR games has garnered interest amongst educational technology researchers. This may be partly as a result of an increase in innovation within the field as well as in the variation in the pedagogical application of VR. However, to the best of our knowledge and evaluation of the existing reviews, no study has attempted to review systematically the trends of EVRGs accounting from the viewpoints of technology, pedagogy, content, knowledge, and games. To bridge this gap, this paper centers in a systematic literature review to examine the current research on EVRGs and to provide important insights beyond the specific research findings within the individual studies.

This phase focuses on the initial preparation undertaken to achieve the goal of this literature review.

Based on the guidelines by Kitchenham and Charters (2007), we identified previous systematic reviews that addressed either our research questions or similar questions. However, as discussed in the introduction, none of the reviews focused on EVRGs. Thus, we decided to conduct a review on EVRGs with specific attention to attributes such as technology, pedagogy and gaming.

According to Kitchenham and Charters (2007), a review protocol provides the basis to carry out the systematic review. Creating a review protocol beforehand helps to reduce the likelihood of research bias, such as potential prejudiced selections of particular studies carried out by the researcher. Formal and informal searches were used to find relevant studies in response to the research question. Table 2 presents the databases and the search strategy of the preliminary searches of existing related studies of EVRGs in education, which set the direction for this study. The databases were selected because they were either specialized in educational technology, educational games or in virtual reality applications, or that have published at least one special issue in educational virtual reality and games for learning. Thereafter, the preliminary review studies presented in Table 1 were used to create the research framework as well as the research question, which yielded a well-organized review protocol.

The second phase of the guidelines by Kitchenham and Charters (2007) follows five stages: the search strategy, the study selection criteria, the study quality assessment, the data extraction plan, and the data analysis tool.

The research question of this study guided the formulation and expounding of the search strategy. As a first step, the search keywords were identified as a way of narrowing down and focusing on relevant articles for the topic under study. The search keywords were chosen according to the research theme; research question and the objective of this study (see Table 2 for search strategy and list of databases). Search keywords: The search space was narrowed down using Boolean search phrases and different combinations of the following terms: “virtual reality” AND “education”, OR “learning”, OR “game”, OR “gamification”, OR “serious game”. While many researchers use the term “educational games” to refer to games, the purpose of which is helping a learner acquire a skill, others opt for a more generic term like “serious game”. The keyword choice was made in a way to capture any game that is used in an educational context, be it formal or informal, as long as it uses VR technology, with the definition we adopted above. The time frame for publication considered within the systematic review was between 2012 and 2018. Since the field of educational technology and virtual reality games is developing very fast, articles before 2012 have been reviewed in previous studies and are not particularly relevant in this work. The screening was based on titles, abstracts, and full-text skimming and took place from 1 February 2019 to 30 May 2019.

Using the keywords and search strings presented in the section “Phase 2.1. Search strategy”, several articles were selected based on their relevance to the research question and inclusion criteria. A flow diagram representing the steps of the selection criteria is presented in Fig. 2. The researchers read the title, abstract, keywords and skimmed through the contents of all the papers and selected the articles that appeared to be appropriate based on VR, education, and gamification. Abstracts, posters, books, and articles that did not show an implementation that required a VR headset were excluded. Altogether, we found 162 research papers through the predefined search keywords on the four databases. However, after we applied the inclusion and exclusion criteria (Table 3), only 31 articles made it through the scrutiny (Table 4).

We set up a coding scheme to guide the extraction of relevant data from research articles with relevance to our research question. The coding scheme included the following overarching dimensions: Title/name of the game, description of the game, country of game implementation, player role, theme, mode, gameplay, goal, platform, VR headset, game interaction, focus, learners, educational setting, research, and evaluation methods.

In summary, the methodology section, presented the planning and implementation process of the systematic review and these included; providing the rationale for choosing the framework adopted for the review, specifying the research questions, developing the review protocol, executing the review, identifying the keywords, drafting the study selection criteria as well as data extraction and analysis. Subsequently, the results yielded by these processes are presented in the next section.

The number of EVRGs publications appearing in each year from 2012 to 2018 is presented in Fig. 3. The bar chart shows a definite increase in the number of research works dedicated to VR educational games. Research shows that there was one study in 2014 and increases to 14 studies in 2018. Table 5 shows that EVRGs have been researched in North and South America, Europe, Asia, and Australia. See Table 5 for full details of serious games.

The review looked at the technological characteristics of the selected EVRGs in terms of platform, headset, and interaction components. As shown in Fig. 4, the Oculus Rift seems to dominate the VR headsets used in EVRGs, as it was used in nearly half of the games (45.2%). HTC Vive came in second used in almost a quarter of the games (22.6%). Cardboard, which is the cheapest of the VR headsets, was only used in 9.7% of the games, while under a quarter (22.6%) of the papers we reviewed, did not specify which headsets were used as target technology.

The platforms used for the games were consistent with the headsets, given that HTC Vive and Oculus Rift DK2 are PC VR devices, 83.9% of the games were playable using a PC or MAC, while 12.9% were playable using a mobile device, which matches the Cardboard requirements (Fig. 5).

In Table 6 (technological characteristics), we show the interaction techniques and hardware used in the EVRGs. While the most basic headsets (e.g. Cardboard) provide basic VR interaction mechanisms like gaze and head movements, more advanced headsets come with more sophisticated controllers. These controllers allow the users to interact with the props in the game environment and facilitate movement. Some games provide natural user interfaces, while other task-specific games provide more advanced controllers such as steering wheels. However, it is noteworthy that traditional input devices, such as a mouse or keyboard are challenging to use with VR technology, as the latter conceals the user in a virtual environment, disallowing the visibility of the surrounding real environment. Some games have, however, suggested such input devices, which makes the gaming experience slightly inconvenient.

The analysis of the games showed that more than two-thirds of the educational VR games were developed for informal learning context (Fig. 6). While the topics taught in the analyzed games were of great variety, we tried to group them under specific themes, to understand which of the tasks or learning goals were deemed VR-appropriate by the researchers. For instance, one-third of the games targeted teaching healthcare-related topics, while a quarter of the games aimed at introducing, training, or enforcing safety measures in various environments (e.g., construction sites, hospitals, or roads). Topics such as biology, physics or astronomy represent a tenth of the surveyed games (represented in Fig. 7 as natural sciences). While topics like language learning, geography, and civil engineering, received minimum interest with one VR game each.

The target audience of the EVRGs depended on the desired learning outcomes. Additionally, the VR games were developed for all levels of formal education, from K-12 to tertiary education students, as well as for lifelong learners who need to acquire some skills to deal with specific health conditions, as illustrated in Table 7.

The 31 EVRGs reviewed in this paper differ significantly in the player roles suggested to their users but are mainly dependent on the educational purpose of the tool. Most of the games suggest player roles that are adequate for the tasks the user is being prepared for, in environments similar to those that they will be working on (see Table 8 for several aspects of gaming characteristics). This shows the power of VR in providing learners with a preview of working activities and conditions.

VR facilitates collaborative learning, where many learners can be together in the same virtual world to execute some learning tasks. Only five EVRGs incorporated such a feature in their designs, while 26 games decided to allow only single-player mode. Gameplay-wise, most games suggest the user moves through the virtual world and executes tasks similar to the real-world tasks they are being prepared for, and so games genres like puzzles are quasi-inexistent in the game set that we have reviewed. This study has found that the VR games present the content to be taught as a series of entertaining challenges to the learner in a virtual environment.

The study revealed that the number of articles or literature on emerging VR systems has been increasing since 2012, which indicates the interest VR has gained since its spread in late 2012. The study also revealed many educational topics in which VR has already been applied. A general emphasis was placed on healthcare education by VR researchers, along with a variety of other topics such as biology, computer science, astronomy, and fire training (Sárkány, 2016; Süncksen et al., 2018). VR has the potential to allow users to experience environments that are otherwise inaccessible in a very realistic way. It allows training in environments that would otherwise be hazardous for learners to train in, as is the case for fire training (Diez et al., 2016). It also allows learners to simulate training with expensive hardware on a risk-free yet realistic environment. Despite these affordances and the application of VR unraveled by this study, there appears to be a lack of research on the use of VR for language learning, failing to benefit from the technology in simulating social interactions, which are very efficient in helping learners to practice their language skills effectively.

Another key finding of this research is the absence of any VR educational games on the African continent. While North and South America, Europe, Asia and Australia participated in developing EVRGs. The cost of technology seems to affect African countries from benefiting from this promising technology. One solution is to rely on affordable headsets supported by Google Cardboard (Bouali et al., 2019). Still, such headsets suffer from the lack of appropriate interaction hardware that allows users to traverse and interact with the virtual world using only gaze and head movement. Such limitation causes Cardboard-related research to engineer games that require minimum interaction while preserving the targeted learning outcomes.

Despite Cardboard being the cheapest VR device, it is the least used technology in the surveyed literature, ranking even lower than the expensive HTC Vive. This raises some questions on the level of adoption of some educational games in real educational contexts, which is not the focus of our study. Nevertheless, the study shows that various games have adopted the use of the Oculus Rift VR, despite being more expensive than Cardboard headsets. Consequently, headsets are not the only problem in VR adoption. This study reveals that interaction is also an issue. In our analysis, we found that most of the games rely on gaze as an interaction mechanism, but this is hardly ever complemented with a natural user interface or interfaces, which facilitate interactions in VR worlds, resulting in the difficulty of using other input devices like keyboards.

Games that target teaching specific skills, such as diving or driving, require more advanced input devices, like Kinect to input body movement or a steering wheel to provide a life-like controller to a vehicle in the virtual world (Calvi et al., 2017; Likitweerawong & Palee, 2018). However, this incurs that such technology can only be afforded by a handful of users, presenting a stumbling block in the adoption of VR games in real-world contexts.

Regarding gaming, most of the gaming environments and contexts were dependent on the real-world context for which the learner is being prepared. This helps in the learning process of the user as it provides him/her with tailored environments that mimic the real world to a higher level of detail. The study also reveals that most of the games developed, allowed only for a single-player mode, failing to benefit from VR’s ability to connect learners in virtual worlds. The dominant game genre in the literature is Role Playing Game (RPG), which is adequate given that the learner impersonates the roles they are being trained for in the virtual world.

The concept of applying digital technology to build a virtual, physical/virtual or hybrid learning environment in which a student experiences a form of play, has emerged over the years and is now gaining acceptance. The puzzle here is that despite the wide variety of VR games developed for different fields (education, healthcare, business, climate, energy, industrial, and financial) only a handful of the games focused on education. Thus, begging the question of why EVRGs have not been widely adopted in mainstream education? Considering the interactive, immersive, and multi-sensory nature of VR, coupled with its growing popularity amongst researchers in psychology, aviation, and cognitive neuroscience, we had expected that even more EVRGs could have been developed and that the technology should have been widely adopted (Alexander et al., 2019). However, there are many reasons for the lack of adoption of EVRGs in mainstream education. One reason is that people tend not to consider VR as a mainstream technology. They perceived that the hype around the technology would lose popularity and be replaced by the reality of the time, in what was called the ‘trough of disillusion’ (Linden & Fenn, 2003). Additionally, there seems to be a lack of know-how by learning technologists and experts on how to design learning solutions with VR. Furthermore, the costs of implementing EVRG for large-scale adoption across educational curricula constitute a limiting factor for adoption in mainstream education (Alexander et al., 2019) and health-related issues.

Christou (2010) states that the three categories of formal and informal educational application areas of VR are used to improve core curriculum subjects such as the applications for edutainment, demonstrations, cultural heritage and museum experiences, as well as an application for training. Our analysis of the pedagogical contributions of EVRGs indicated a considerable variation among the subjects that were implemented. For example, the trend of EVRGs in this study showed that 32% of the applications were developed to support medical education. In contrast, only 3% was implemented to support subjects that involve the concretization of abstract concepts such as physics and geography.

Although the learners in these studies balanced well amongst children, youth and adult learners, there are obvious variations amongst the type of learners about the kind of EVRGs. For example, most learners could have medical conditions such as patients with neck or back pain, dementia, autism spectrum disorders, post-stroke conditions, brain lesions, and visual impairment. Translating clinical procedure into EVRGs to facilitate learning and future implementation of the procedure by caregivers and patients constitute the main interest in developing medical EVRGs (Heuven et al., 2017; Mihajlovic et al., 2018). Simulating the medical procedure offer motivation, engagement, and positive learning experience among EVRGs users (Trombetta et al., 2017). Having examined an essential pedagogical attribute, of the context of the learner in this study, we noticed that the learning context for most EVRGs falls within the informal settings. As several EVRGs are developed to provide particular training solutions to the user, the informality of the setting tends to overshadow the necessity of a classroom environment (formal setting).

The limitations of this study include: