Visiting real environments is not always an option for human-environment interaction research and surrogate exposure methods (e.g., slideshows) are tricky to manipulate (e.g., weather, or physical structure) and do not easily foster immersion. Conversely, virtual reality (VR) is becoming more accessible and enables a bespoke, dynamic, and immersive experience. As a result, there is growing interest in VR as a research tool and several studies have leveraged its ability to create controlled experimental settings. However, further investigation is needed to understand the limits and appropriate applications of this technology in research.

Through three studies this thesis explores the past, present, and future of VR technology in human-environment interaction research, providing recommendations to ensure the continued production of high-quality data.

For Study One a total of 16 student and non-student participants from the south of England were recruited (n_female = 8). Using a within-subject design this hybrid field-laboratory experiment examined whether VR could sit alongside contemporary methods by exploring perceptions of the same lake presented in the real world, in VR and as a video. Self-reported mood and enjoyment recorded at baseline and after each ten-minute exposure showed that experiences of VR presentations fell between real and video presentations. Additionally, in a short survey with open-ended questions, participants stressed the importance of naturalistic interaction, sensory immersion, and graphical realism in the experiences, frequently using pleasure and contentment words, with VR also eliciting excitement words. Results showed a preference for real and virtual experiences over watching a video. However, real and virtual experiences were not found to be particularly different to one another, except for comments regarding graphical realism. This calls into question the cost-benefit of producing hyper-realistic experiences, especially in relation to participant’s perceptions of and expectations regarding surrogate nature.

For Study Two a total of 120 student and non-student participants from the south of England were recruited (n_female = 101). Using a between-subjects design, this laboratory experiment examined how affective responses are impacted when the level of realism of natural and built scenes are altered in VR. Stress was induced via a writing task before participants spent ten minutes in a high or low realism version of a natural or built environment. Participants self-reports of mood and sense of presence over time showed that more realistic VR environments provided greater recovery from stress and a greater sense of presence. As with Study One participant review of the experience stressed the importance of naturalistic interaction, sensory immersion, and graphical realism in VR experiences. Results show that while low realism environments do yield viable results, stronger effects are produced by higher realism environments. Thus, researchers using VR should endeavour to produce more realistic experiences, both in terms of graphical fidelity and naturalistic interaction. Considering this, what can be learned from existing researcher experience in order to help the community produce optimum outputs when working with this relatively new technology.

For Study Three a total of 102 researchers who have published, in the past five years, VR research that studied human-environment interaction from Asia, Europe, and America were invited to participate. Valid responses were collected from nine researchers (n_female = 4) at various stages in their academic career. An online questionnaire comprised of open-ended and multiple-choice questions was designed to ascertain the future of VR as a research tool. The questionnaire explored personal experience of working with the technology. A thematic analysis of the responses revealed advantages relating to recruitment, multidisciplinary collaboration, and facilitating research. Identified disadvantages were related to health and safety and cutting-edge limitations. Future engagement with VR technology in human-environment interaction research was identified to be dependent on technological advancements, health and safety, additional research, and multidisciplinary collaboration. An appraisal of VRs ability to facilitate the collection of good quality data was favourable with its performance in terms of validity, reliability, and generalisability receiving slightly effective ratings. Clearly VR is a tool useful in human-environment interaction research, overcoming issues relating to experimental control and facilitating innovative new strands of research. However, there is still work to be done by the researchers and technical experts to mitigate health and safety risks, whilst also bridging the knowledge gap between the disciplines. As VR continues to evolve it is important that future research that attempts to leverage these benefits maintains awareness of the potential drawbacks and what can be done to mitigate their influence.

While real experiences remain the optimum choice for studying our relationship with an environment, this work shows that VR offers a suitable replacement and has great potential for advancing scientific enquiry by enabling environmental control, manipulation, and accessibility. However, consideration must be given to how participants’ perceptions and expectations regarding VR impact research. Additionally, when using VR where a visual component is necessary researchers should endeavour to produce more realistic experiences, both in terms of graphical fidelity and naturalistic interaction. Finally, while the onus for the future of the technology as a research tool lies in the hands of the developers, researchers should continue to collaborate closely with those in this discipline to address health and safety risks, as well as knowledge gaps.