Introduction
One of our aims with this book is to move beyond the idea that research with VR can only take place via a complex process of content creation. This and the following chapter, therefore, focus on commercially available, ready-made materials. The use of existing materials lowers the barrier of entry for researchers wanting to work with VR, and there are significant research gaps that can be addressed through engaging with this content.
There are, of course, non-commercial existing VR experiences that can be reused in research projects as well. These can, however, sometimes need a bit of detective work in order to track down the owner of the material, find out whether they are willing to share it and, crucially for older content, whether it can still be made to function on available hardware. Permissions would also be needed for it to be employed in a different context from its design purpose, which may require careful consideration. The reconstruction of the Auschwitz concentration camp that we highlight in Chapter 6, for example, might potentially form the basis of a new project
Commercially available VR materials, meanwhile, can generally be reused more freely in a variety of different research contexts, though may need to be purchased before use. Such products can also suffer the same technical obsolescence that we see in non-commercial VR content, with the risk of becoming abandonware, where the developer sees no value in maintaining the material. Commercially produced material does, however, often have the advantages of scale and quality control that comes through professional development, as well as being relatively simple to find and install via online platforms such as Steam and the Oculus store.
Games are the most common form of commercial VR content, but there is also edutainment material and productivity and well-being apps, as well as more specialised industrial and medical software. In this chapter, we examine ways to undertake content analysis with existing VR materials; this represents quite an important research gap within the field. While a solo researcher undertaking a close textual analysis is a bread-and-butter approach in the humanities and has been widely employed within game studies, VR materials have largely escaped this kind of critical lens. Another significant research gap is in undertaking a broad survey of a particular VR genre – well-being apps or medical trainers for example – to examine the kinds of discourses being employed by developers and how these have changed over time.
There are challenges when employing a content-analysis type of approach to VR materials. Many of the analysis techniques employed within game studies and other disciplines are directly transferable, but need to be augmented with techniques for examining bodily movement. We will consider some of these issues in turn before moving on to a worked example where we examine the practicalities of undertaking an analysis of the VR game Half-Life: Alyx (Valve, 2020).
Much of the work around existing VR materials focuses on the user experience and how participants respond to the content – issues that we will explore in Chapter 3. There has, however, been relatively little use of content analysis approaches by scholars working on VR. Where it does exist, these studies primarily take the form of content analysis of materials about VR rather than of the VR experiences themselves. Nevertheless, the second approach has produced some very interesting work. Keller et al (2017), for example, examined how a video about the use of VR in healthcare was received by social media users. Here, a quantitative analysis of 2,401 Facebook posts demonstrated a generally positive perception of the technology, with women slightly more favourable than men. This kind of study is useful for assessing public concerns about the use of VR technology and how these might be overcome, which is particularly significant in a healthcare context. Johnston et al (2017), meanwhile, note that, while VR has created a great deal of excitement within the education sector, the pedagogies underlying its use are not clearly articulated. They used a purposive sampling technique to examine publicly available websites and video material promoting and discussing educational VR content, alongside a systematic review of relevant academic literature. Of the 35 pieces of educational software they examined, the overwhelming majority (24) focused on experiential learning approaches. This is perhaps unsurprising given that experiential learning theory emphasises experience and interaction as key modes of learning, which has obvious resonances with the immersive and embodied qualities of VR.
Studies of these kinds operate at one remove from the actual VR experience itself, however. Johnston et al (2017) were investigating published material about different pieces of educational VR software, rather than examining that software directly and attempting to code its content. There is a notable
Mindfulness discourses tend to position the individual as being responsible for their own self-improvement, rather than acknowledging the structural issues within a neoliberal economy that act as a barrier to this (Pykett and Enright, 2015). Given the emphasis on experiential learning approaches within VR, examining how these discourses of self-improvement are actualised by VR apps would be a very interesting research project. There is, however, a bit of a ticking clock here, as some of the apps developed during the initial excitement around the third wave of VR from around 2012 are themselves starting to become obsolete and may soon become inaccessible. An example of this is the range of #befearless apps developed by Samsung for its phone-based Gear VR headset. These apps were designed to provide users with virtual experiences, such as giving practice presentations in front of an avatar audience or ‘walking’ across high bridges, suggesting these could help people conquer their fears. Support for the Gear VR headset was, however, withdrawn by Samsung and the device rendered non-functional by software updates in 2020. At the time of writing, the #befearless apps are still available for the Oculus Go headset, but this device is itself due to have its support removed, meaning that these apps may become permanently inaccessible.
There is merit to researchers simply documenting those VR experiences that are at risk of becoming abandonware. This would provide material for subsequent content analysis of how different VR genres evolve over time and the kinds of discourse embedded within them. Such an approach has commonalities with Miller and Garcia’s (2019) work on ‘digital ruins’. They examined online 3D worlds that, having fallen out of fashion, no longer had a significant user base to animate their virtual spaces. Methodologically, this project employed an autoethnographic approach, with the researchers visiting these semi-abandoned worlds and documenting their experiences. Again, this process of documentation is significant not least because Blue Mars, one of the three sites they investigated, has since been shut down. The relatively straightforward processes of capturing stills and videos from the researchers’ interactions alongside taking notes of thoughts and reflections are key to documenting this kind of content.
Different forms of content analysis are commonly used within the field of game studies and allied research. As a result, work in this area offers useful methodological insights for approaching VR material. Game studies as a discipline owes some of its origins to media studies and although it takes an interdisciplinary and integrative approach, many of its methods and techniques are familiar to scholars within the broader humanities. Lankoski and Björk (2015) have compiled a useful collection that explores how conventional methods from the humanities and social sciences can be adapted to meet the specific challenges of working with games, from qualitative content analysis to much more complex modelling approaches. A key reason why conventional methods need to be adapted is that, unlike other forms of media, such as books or photographs, one of the unique qualities of games is their interactivity. This issue was codified at an early point in the development of game studies as a tension between narratology,
Although game studies is a methodologically diverse field, approaches to content analysis here tend to fall into one of two camps: reviewing multiple games to analyse a particular subgenre or theme (for example gaming nostalgia explored by Sloan, 2016), or close textual analysis of a single or small number of games (such as discourses of disability in Mass Effect, Jerreat-Poole, 2020). Both approaches have relevance to scholars interested in working with existing VR content. A review at the subgenre scale can be used to explore how discourses shift over time (such as the evolution of first-person shooters (FPSs), Hitchens, 2011). This can be particularly valuable when examining eras with the kind of rapid development we are seeing in the third wave of VR. Close textual analysis, meanwhile, is very familiar among literary scholars and others in the humanities and social sciences, with a variety of approaches taken. Felczak (2020), for example, offers a nice illustration of the specific challenges faced when undertaking this type of analysis within games. This paper undertakes a postcolonial examination of Pillars of Eternity 2: Deadfire (Obsidian Entertainment, 2018), noting that the designers were clearly telling a story that highlighted the exploitations and violence of the colonial period. The implicit decolonial critique of the game’s story, however, can be seen as coming into tension with the gameplay and its ‘heroic, power fantasy tropes’ (Felczak, 2020) typical of role-playing games. The value of Felczak’s analysis is in exploring how these elements are enmeshed in producing the gaming text.
Analysing embodied engagement
In studying games, one cannot meaningfully separate story and gameplay. The immersive quality generated by HMDs make this interplay of content and interaction even more acute by generating the sense that the user is inside the virtual environment. Any content analysis of VR materials therefore needs also to consider the body of the user. Questions around embodiment have been of particular interest to geographers over the last two decades, with many projects examining bodily interactions with different spaces partly in response to an influential canon of theoretical work around the performative and non-representational (for example Thrift and Dewsbury, 2000). While these theoretical positions can seem somewhat opaque and insular to outsiders, they have helped to stimulate some really interesting practical development of methodologies for exploring how our bodies shape an understanding of the world around us. Wylie (2005), for example, has used autoethnographic approaches to
Although the body plays a less prominent role in the experience of conventional gaming, considerations of the bodily are not absent from conventional game studies, with useful methodological lessons to draw upon. There has been a great deal of interest in how game designers attempt to maximise players’ ‘flow’, drawing on Csikszentmihalyi’s (1990) ideas of creating a psychological state where the individual has an optimised experience. Flow has been used to think about the ways in which games seek to maintain player engagement and enjoyment so that they play for longer and are more willing to buy new gaming products. The maintenance of flow requires a consideration of the bodily, from controllers that players can use to seamlessly control the on-screen action (Schmalzer, 2020) to sound design that creates affective responses in players (Oldenburg, 2013).
Considerations of the bodily are very significant within VR research, but primarily from the perspective of examining the physical effects of using these systems. We will explore the problem of cybersickness in more detail in the next chapter, but, at a basic level, it partly comes from the mismatch between bodily movement and position on screen. Indeed, it can even occur where expensive systems of treadmills and specialist footwear are employed to create a sense of physical movement within a constrained space (Wehden et al, 2021). Beyond cybersickness, however, as we go on to explore, physical considerations can have a major impact on the practicalities of undertaking research in this area.
Case study: approaching a content analysis of Half-Life: Alyx
In order to examine how these more abstract considerations play out in practice, we turn now to consider a worked
Situating the content
In writing up any content analysis, there is a balance to strike between being overly descriptive and giving the reader sufficient contextual material that they can understand the point being made without having directly engaged with the text themselves. This is a particular issue within game studies, where some game texts are part of a long-running series with a sprawling and complex lore. As such, there can be a tension between situating a game within a wider narrative arc or treating it independently, both in terms of writing it up, but also in approaching the text in the first place.
Although Alyx is part of a long-running franchise, we chose to treat it independently and deliberately did not play the earlier games or read up on the wider story of which it is a part, as we wanted to approach it with a clean slate. There are disadvantages to this approach in that many references to parts of the wider story may be lost, but the advantage is in exploring the text without preconceptions. A good compromise between these positions is subsequently to learn more about the wider story and play the other games in the series before undertaking a second documented playthrough of the main text being studied.
In terms of the story, the wider Half-Life franchise imagines a post-apocalyptic future following an alien invasion. Players in the earlier games inhabit the main protagonist, physicist Gordon Freeman, from a first-person perspective and have to solve various puzzles, run around, drive different vehicles and shoot seemingly endless waves of different aliens and soldiers in a quest to liberate humanity. Highly innovative when the original game was released in 1998, Half-Life and its sequels have been imitated to the point of seeming not just technically dated but also somewhat derivative. Valve have publicly stated that they only wanted to return to the franchise once they could create something as technically innovative as the first game, and they saw VR as the opportunity to do this (Wilde, 2020).
Alyx takes place in the same world as the earlier games, but the player inhabits Alyx Vance, a young woman of Afro-Asian descent who previously appeared in the franchise as a non-playable minor character. The basic storyline of Alyx is not dissimilar to the earlier games, with players tasked with surviving attacks, exploring and solving puzzles alongside
An advantage of researching a major game such as Alyx is that developers are often interviewed by journalists about the kinds of choices they made in putting the game together, which can add value to the content analysis. Because Alyx is so innovative and unusual, there is a wealth of secondary material that can be consulted in which the developers have reflected on the process of game design, particularly in the pacing of the story so as not to overwhelm the player with too many stimuli before they are used to being in the virtual environment. Indeed, some of the intended gameplay had to be slowed down to take account of how intense the VR experience is: during testing, some of the faster-moving creatures from the earlier games were removed because players found that they simply could not cope (Rad, 2020). Beyond secondary sources, developers are often open to being interviewed directly by researchers seeking more specific information about design decisions and, again, this is a well-established technique within game studies (for instance Haylot and Wesp, 2009) that could also be applied to VR content.
Documenting the experience
At its simplest, content analysis of games requires the researcher to play through the material – potentially multiple times – while taking notes about plot and gameplay alongside the visual and auditory stimulation and researchers’ reflections upon the experience (Jones and Osborne, 2021). Given that the main story arc of a triple-A title can take upwards of 20 hours to complete, this is not an insignificant investment in time for the researcher. A non-gaming VR experience might
For the Alyx project, we set up a play space of approximately 2m by 3m and positioned a video camera on the edge of this area, capturing not only the player’s movement but also a large TV screen mirroring the output from the HMD. One of the most powerful qualities of VR is in making imagery that looks unremarkable on a conventional monitor feel incredibly compelling in an HMD, even where the graphics are not particularly sophisticated. Alyx, with its triple-A production values, feels shockingly real in VR, though when looking at footage captured via a mirrored output, it appears to be simply a good-quality modern game. This emphasises the importance of considering the captured video material purely as a set of notes or aides-memoires for the researcher, given how different this imagery is to the VR experience itself. The game itself was played using a Razer Blade Pro laptop (7th gen Core i7, GTX1060 graphics card) in combination with a tethered Samsung Odyssey HMD.
The first phase of data collection was a collaboration between Phil and Tess, with one person playing and the other asking the player questions to prompt reflections on what was being experienced at different moments. The person not wearing the HMD also played a practical role of making sure that the player did not trip up or accidentally collide with objects in the physical world, given how quickly someone in VR loses any sense of having a body that exists outside the virtual space. This also had the unanticipated advantage of providing an anchor into a safe space, which proved highly valuable given the horrifying nature of much of the experience within Alyx.
Tess:I’m not going down there. I can’t.Phil:OK.Tess:I physically can’t. I feel sick thinking of going down there.Phil:OK.Tess:That scares me far too much. … Oh, look at how nasty that looks. I’m sorry I’m so pathetic, but every inch of me is saying, ‘No, do not go down there.’(Extract from video recording, 22 July 2020)
Of course, this highlights questions about the extent to which VR experiences can exceed the player’s embodied, affectual limit, with implications both for who is able to undertake
Peering down into a dark sewer tunnel in Half-Life: Alyx
Source: Phil Jones and Tess OsbornePeering down into a dark sewer tunnel in Half-Life: Alyx
Source: Phil Jones and Tess OsbornePeering down into a dark sewer tunnel in Half-Life: Alyx
Source: Phil Jones and Tess OsborneIn conventional games, the player needs a period of adjustment, learning how different buttons on the controller relate to different actions on screen. The same is true in Alyx, but with an added layer of complexity because there are specific bodily movements to learn in order to accomplish particular tasks. Reloading a weapon, for example, requires a combination of button presses on the hand-tracking controller and physical movements – reaching over the shoulder to collect a magazine from a virtual backpack, bringing hands together in a particular way to put that magazine into the gun. Some of the weapons also need a two-handed manoeuvre to cock the gun. The video footage captures some of the learning process of becoming more practised and smoother at reloading weapons during – often quite hectic – combat sequences. Indeed, in notes recorded later in the playthrough, Phil talked about this becoming quite a physically satisfying process. To begin with, however, the unfamiliar set of motions created moments of abject panic, with a scurrying headcrab jumping at his face while he screamed, struggling to remember how to reload the weapon that he had emptied by firing wildly, desperate to stop the alien attack. We did not monitor heart rate or electrodermal activation as we have on other gaming projects (Osborne and Jones, 2017), but it is clear that both would have seen a dramatic spike in these moments. Indeed, the horror combat sequences fighting in the dark left both of us physically and emotionally drained and needing a break from being inside the game.
The data collection took place in the summer of 2020, which was unusually hot for the UK. Wearing a heavy, sweaty HMD while moving around, with game sequences creating a very high heart rate, the VR activity was simply tiring and we tended to play in 20–30-minute bursts. The combination of the heat and the stress left both of us feeling headachy and nauseated
Analysing the materials
At the end of the data collection with a full playthrough completed, we had recorded approximately 21 hours of footage over 13 days spread across two months. Audio from the video recordings was transcribed using an automated process with manual correction, dividing between commentary made during the gameplay itself and our more reflective discussions delivered to camera immediately after coming out of VR. These recordings also captured the gameplay and story elements that were displayed via the secondary TV output, though these were not transcribed. The more narrative elements could have potentially been captured to a higher quality by using screen recording, but we avoided this for technical reasons, concerned that Alyx was already pushing the limits of a laptop at the bottom of the specification needed for the game. Thus, while the basics of the story were captured as a background
With transcribed notes of conversations, responses during gameplay and immediate reflections alongside footage of player movement and on-screen action, we had a wealth of material to examine. Different disciplines have different preferred approaches to content analysis, with different levels of complexity, depending on what the researcher wants to achieve. Because we were simply exploring the practicalities of how a VR experience might be examined through this lens, we have not as yet undertaken a major analysis of this dataset beyond some simple thematic coding. Even with a relatively superficial examination of the dataset, however, the importance of physicality to the experience came through very strongly: if we had examined the transcript data in isolation, some quite significant elements about this would have been missed. Thus, there is an imperative to draw on some of the techniques developed for analysing video footage within research projects to situate and make sense of some of the comments captured in the transcripts.
There are a number of useful techniques here. The transcripts acted as a jog to the memory, to recall particularly significant sequences that could then be examined in more detail via the video material. This allowed for descriptions of sets of movements to sit alongside the transcript material. There have been some interesting projects for analysing video that have made use of a kind of comic-strip approach to capturing a sequence of events through a series of stills and which could be useful when considering how to represent this kind of material for publication (Lloyd, 2019). We have also experimented with making animated gifs for key sequences to use in presentations, capturing five- or ten-second moments that illustrate a particular point. While the gifs are of limited use for conventional publication, they can form useful supplementary material where publishers offer this, or can be embedded within articles where publishers operate an online-only model.
Conclusion
The horror genre is built around unsettling and discomforting its audience, while maintaining a safe distance from the events being depicted. The sense of dislocation within horror works well for VR experiences, although being placed within the events can be quite distressing despite the knowledge that all one has to do to escape is remove the HMD. This highlights how VR experiences are qualitatively different from conventional media such as films, TV and games. Thus, while we can draw on approaches to content analysis from media or game studies, any examination of VR content needs to go further to bring questions of embodiment to the fore. As we have illustrated, this can be quite an intense and complex process that asks questions of how the material body intersects with the virtual space.
As we will see throughout the rest of this book, much of the work on VR comes from an explicitly scientific position where there is not the disciplinary tradition of using content
From a scholarly point of view, there is also a risk that parts of the history about the developing discourses embedded in VR experiences are going to be lost as earlier pieces of software and hardware become obsolete and non-functional. As a result, there is real value to applying content analysis approaches to these materials. Alyx is an unusual piece of software because of its scale and ambition, but many VR experiences are considerably smaller and would not require such a large commitment of time to examine. Thus, for example, an examination of VR well-being experiences as a parallel to Mani et al’s (2015) work with iPhone apps could be fairly straightforward to undertake and would give some very interesting insights into the difference that being embodied within the experience makes to the way these discourses are shaped.
References
Bates C (2013) Video diaries: audio-visual research methods and the elusive body. Visual Studies 28(1): 29–37.
Csikszentmihalyi M (1990) Flow: The Psychology of Optimal Experience .New York: Harper & Row.
Felczak M (2020) Dungeon pirates of the postcolonial seas. Domination, necropolitics, subsumption and critical play in Pillars of Eternity 2: Deadfire. Game Studies 20(4).
Frasca G (2003) Ludologists love stories, too: notes from a debate that never took place, in M Copier and J Raessens (eds) Proceedings of DiGRA 2003: Level up. Utrecht: University of Utrecht, pp 92–99.
Haylot E and Wesp E (2009) Interview with Brad McQuaid and Kevin McPherson. Game Studies 9(1).
Hitchens M (2011) A survey of first-person shooters and their avatars. Game Studies 11(3).
Jerreat-Poole A (2020) Sick, slow, cyborg: crip futurity in Mass Effect. Game Studies 20(1).
Johnston E, Olivas G, Steele P, Smith C and Bailey L (2017) Exploring pedagogical foundations of existing virtual reality educational applications: a content analysis study. Journal of Educational Technology Systems 46(4): 414–439.
Jones P and Osborne T (2021) Gaming and VR in geographical research, in N von Benzon, M Holton, C Wilkinson and S Wilkinson (eds) Creative Methods for Human Geographers. London: Sage, 285–295.
Keller MS, Park HJ, Cunningham ME, Fouladian JE, Chen M and Spiegel BMR (2017) Public perceptions regarding use of virtual reality in health care: a social media content analysis using Facebook. Journal of Medical Internet Research 19(12): e419.
Lankoski P and Björk S (2015) Game Research Methods: An Overview .Pittsburgh, PA: ETC Press.
Laurier E and Philo C (2006) Possible geographies: a passing encounter in a café. Area 38(4): 353–363.
Lloyd M (2019) The non-looks of the mobile world: a video-based study of interactional adaptation in cycle-lanes. Mobilities 14(4): 500–523.
Mani M, Kavanagh DJ, Hides L and Stoyanov SR (2015) Review and evaluation of mindfulness-based iPhone apps. JMIR mHealth and uHealth 3(3): e82.
Miller V and Garcia GC (2019) Digital ruins. Cultural Geographies 26(4): 435–454.
Obsidian Entertainment (2018) Pillars of Eternity 2: Deadfire. Baltimore, MD: Versus Evil.
Oldenburg A (2013) Sonic mechanics: audio as gameplay. Game Studies 13(1).
Osborne T and Jones PI (2017) Biosensing and geography: a mixed methods approach. Applied Geography 87: 160–169.
Pykett J and Enright B (2015) Geographies of brain culture: optimism and optimisation in workplace training programmes. Cultural Geographies 23(1): 51–68.
Rad L (2020) Half-Life: Alyx cut enemies that were too scary, inVR, available at: https://www.gamespot.com/articles/half-life-alyx-cut-enemies-that-were-too-scary-in-/1100-6475051/ [Accessed 7 April 2021].
Schmalzer MD (2020) Janky controls and embodied play: disrupting the cybernetic gameplay circuit. Game Studies 20(3).
Sloan RJS (2016) Nostalgia videogames as playable game criticism. GAME: The Italian Journal of Game Studies 5.
Thrift N and Dewsbury J-D (2000) Dead geographies – and how to make them live. Environment and Planning D: Society and Space 18: 411–432.
Valve (2020) Half-Life: Alyx. Bellevue (WA): Valve Corporation.
Wehden L-O, Reer F, Janzik R, Tang WY and Quandt T (2021) The slippery path to total presence: how omnidirectional virtual reality treadmills influence the gaming experience. Media and Communication 9(1): 5–16.
Wilde T (2020) Behind the scenes of ‘Half Life: Alyx’: how Valve revived a classic franchise for the VR era, available at: https://www.geekwire.com/2020/behind-scenes-half-life-alyx-valve-revived-classic-franchise-vr-era/ [Accessed 9 August 2021].
Williams V (2018) Frameless fictions: exploring the compatibility of virtual reality and the horror genre. Refactory: A Journal of Entertainment Media 30.
Wolf & Wood Interactive (2016) A Chair in a Room. Gateshead: Wolf & Wood Interactive.
Wylie J (2005) A single day’s walking: narrating self and landscape on the South West Coast Path. Transactions of the Institute of British Geographers 30(2): 234–247.
