Virtual Reality (VR) and Augmented Reality (AR)

“Immersive” does not describe a single device, teaching method or treatment.

Virtual reality (VR) presents a simulated environment, often replacing the surrounding view through a headset. Augmented reality (AR) adds digital information to a view of the physical world. It can appear on a phone, tablet or head-worn display. Extended reality (XR) is an umbrella term covering these approaches.¹

“Mixed reality” is used inconsistently across products; it often emphasizes spatial anchoring and interaction between digital objects and the surroundings. Seeing the room may involve transparent optics or a camera feed. Neither arrangement guarantees an unobstructed, perfectly accurate view.

Hardware quality matters, but no universal resolution or latency figure describes all headsets. Fit, tracking, readable text, movement design and the application itself can matter more to a particular task than a headline specification. A desktop simulation may also be called “VR” in a research paper, so always check what participants actually used.

Two useful ideas are presence, the feeling of being in a depicted environment, and agency, the sense that one's actions influence events. The research-informed CAMIL model proposes that these features can influence interest, motivation, cognitive load and self-regulation. It is a framework for understanding learning, not proof that immersion automatically improves it.²

Immersion can also compete with the task. A learner may spend attention on moving through a scene, operating controllers or admiring the surroundings. A realistic experience can feel important even when the underlying information is inaccurate or poorly explained.

Consider a virtual historical site. It might help a learner understand scale and geography, but a convincing reconstruction should still identify what is documented, what is inferred and what is artistic interpretation. Visual realism does not remove the need for source literacy.

A meta-analysis of 35 experiments found a small average learning advantage for headset-based immersive VR over less immersive or conventional comparisons. That average covered varied lessons, learners and designs; it is not a promised gain for the next classroom.³

A separate systematic review highlighted recurring limitations: short interventions, limited assessment of retention and inconsistent outcome measures. Both reviews cover an earlier generation of studies, but their central evaluation questions remain useful.⁴

Ask what the comparison really was

Extra practice, better feedback, a more carefully designed lesson or additional teacher attention can improve results independently of immersion. Compare an XR activity with a credible alternative, giving learners similar time and support. An engaging, well-designed desktop lesson is a more informative comparator than doing nothing.

Good teaching does not begin when the headset goes on. In a randomized media-and-methods experiment, preparatory instruction improved immediate knowledge and transfer scores in the immersive condition. Giving learners a mental map can help them spend less effort figuring out what the environment is showing.⁸

Another experiment improved learning by combining segmented VR viewing with written summaries. Because the redesign changed both features, it does not prove that either one alone caused the benefit. It does support a practical principle: create opportunities to stop, organize and explain.⁹

1. Before: name the learning goalExplain the relevant concepts and controls. Show what success looks like, how to pause and which non-headset option is available.
2. During: make actions purposefulAsk learners to compare, predict, inspect or test something. Remove visual flourishes that compete with the question.
3. Between activities: make sense of itInvite a short explanation, sketch or discussion. Use misunderstandings to guide the next activity rather than treating completion as understanding.
4. After: test beyond the sceneUse a fresh problem or physical task, and return to the idea later. Record both learning outcomes and any difficulty participating.

An example: understanding a rotating machine

Begin with a labeled diagram of the main parts. In the simulation, ask learners to predict what will happen when one part changes speed, then inspect the result. Outside the headset, have them explain the mechanism using a different diagram. A later question can test whether the explanation still holds without the original visual cues.

This is a teaching example, not a validated intervention. Its value lies in making the objective and assessment explicit. The same approach can work with a physical model or a desktop animation; choose XR only if its spatial or interactive features add something useful.

A therapeutic experience needs a clinical rationale, an appropriate participant, a tested protocol and a way to respond when things are not going well.

VR can make some situations easier to arrange and adjust. In treatment for a fear of heights, for example, a person can encounter carefully designed height-related scenarios without beginning at a real high place. The goal is useful learning and increased freedom outside the simulation, not simply enduring discomfort inside it.

Fear of heights: evidence for a specific intervention

A 2018 randomized trial involving 100 adults found substantial reductions in fear after an automated, avatar-guided VR cognitive-therapy program compared with usual care, with benefits still evident at four weeks. The comparison was not with an equally intensive therapist-delivered treatment. It supports this designed intervention; it does not establish that any frightening VR experience is therapeutic.¹⁰

Anxious avoidance: a different clinical target

The gameChange trial studied 346 people with psychosis who had anxiety-related difficulty going outside. Adding supported VR therapy to usual care produced modest average improvements in avoidance and distress at six weeks; whole-sample differences on the primary self-report outcomes were not significant at 26 weeks. The target was anxious avoidance, not a general cure for psychosis.¹¹

Relaxing nature scenes may be enjoyable, but a calming experience is not automatically treatment for an anxiety disorder or trauma-related difficulty. Equally, immersive technology cannot replace supportive relationships or resolve an unsafe real-world situation.

Before clinical use, ask what condition the program addresses, what evidence supports it, who oversees care and what alternatives are available. Participation and stopping should remain meaningful choices, not tests of cooperation.¹²

Different clinical applications use immersion for different reasons. A distraction experience during a procedure is not the same intervention as a multiweek pain-management course. Movement practice after stroke asks different questions again. Combining them into one claim that “VR works” loses important information.

Chronic pain: skills rather than a virtual cure

In 2021, the US FDA authorized marketing of EaseVRx, a specific prescription immersive system providing adjunctive behavioral treatment for adults with chronic moderate-to-severe low-back pain. Its intended outcomes concern pain and interference with daily life—not repairing an underlying spinal condition. This was authorization of a particular medical device, not approval of VR as a whole.¹³

An eight-week randomized study of the therapeutic program found improvements in several pain and interference outcomes compared with sham VR. The sham group also improved. The study involved a selected sample and manufacturer funding, so baseline-to-follow-up percentages should not be presented as the treatment's comparative effect or a promise to every patient.¹⁴

Stroke rehabilitation: what is doing the work?

The 2025 Cochrane review found small benefits for some outcomes when VR was compared with a similar amount of alternative therapy, with uncertainty varying by outcome. Adding VR to usual rehabilitation could improve arm function, but also added practice time. Most included systems were nonimmersive, and few studies tested modern headset-based immersion.¹⁵

AR offers possibilities without fully replacing the physical view. A small trial involving 30 people after stroke reported advantages on selected measures after a two-week cellphone-based AR program with matched total therapy time. That is promising early evidence, not proof of durable recovery or superiority for every rehabilitation need.¹⁶

Home access still requires a care plan

Remote delivery may reduce travel, but someone still needs a suitable space, usable equipment, clear instructions and help when problems arise. Ask which movements are appropriate, how progress is reviewed and when to contact the care team. A consumer fitness or relaxation app should not be assumed interchangeable with prescribed rehabilitation.¹²

If someone is nauseated, strained or worried about falling, the experience is not succeeding simply because the software continues running.

Cybersickness has several contributors

Nausea, dizziness, headache and disorientation can occur during or after immersive use. Mismatches between visual motion and bodily motion are an important explanation, alongside factors involving display behavior, movement design and individual susceptibility. There is no single prevalence figure that applies across all users and applications.¹⁷

A stationary viewpoint, predictable movement and controls that avoid unnecessary simulated acceleration may improve tolerability for some people. None guarantees comfort. A participant who stops early must still count in an evaluation; excluding them can make an application appear easier to use than it really is.

Eyes, focus and fit

In many stereoscopic displays, the eyes converge toward a depicted depth while focusing at the display's optical distance. Research on this vergence–accommodation conflict shows that mismatched depth and focus cues can contribute to visual discomfort. The effect depends on the display and viewing conditions.¹⁸

Eye strain and dryness can occur during prolonged viewing. These symptoms should not be confused with proof of permanent eye damage, but they are still reasons to adjust or stop. Persistent visual symptoms deserve assessment by an eye-care professional.¹⁹

The room still exists

Clear the usable area, check cables and fit, and consider a seated option when appropriate. Headsets are not protective equipment. A boundary graphic or camera view does not remove collision risk, and awkward fit or weight can contribute to discomfort. Follow the specific device's instructions rather than adding unapproved counterweights or accessories.²⁰

People with relevant visual, balance, neurological or musculoskeletal conditions may need individualized advice before use. The right choice depends on the person, the hardware and the content—not age alone.

A 2024 systematic review of headset VR in children under 14 found limited evidence of harms from short, supervised exposures, but also poor and inconsistent reporting of adverse effects. It called for better research on repeated exposure. That is neither proof of long-term safety nor evidence that all childhood VR use is harmful.²²

For children, consider fit, content, supervision and the ability to describe discomfort. Follow current manufacturer age guidance and the relevant account rules; meeting an age requirement does not settle every developmental or safety question. The American Academy of Pediatrics also emphasizes family discussion, appropriate content, breaks and avoiding displacement of other important activities.²³

Accessibility is not one switch

A person may need captions, high contrast, alternatives to precise hand movements, adjustable reach distances or seated interaction. Others may need less sound, a simpler visual scene or a way to operate controls without gaze selection. W3C's XR Accessibility User Requirements describes a range of such needs; it is guidance, not a guarantee that a particular product meets them.²⁴

For example, “look at the button to select it” may be convenient for one person and tiring or unreliable for another. A meaningful alternative input method matters more than a generic claim that the interface is intuitive. Captions also need to remain readable without forcing uncomfortable head movements.

Social safety in a shared space

Real-time voice, gestures and embodied avatars can make unwanted interactions feel particularly intrusive. Australia's eSafety Commissioner recommends accessible exit, mute, block and report controls, alongside personal-boundary settings and safety-by-design measures. Responsibility does not rest solely with the person experiencing harassment.²⁵

Before entering an unfamiliar shared environment, find those controls and decide who can contact you. A personal boundary can limit some interactions, but cannot prevent every form of verbal abuse, recording or unwanted contact. Schools and clinics should know who can enter a session and how an incident will be handled.

Movement can be identifying

A 2023 study involving more than 55,000 VR users showed that head-and-hand motion could identify people within an enrolled dataset after training on their previous recordings. It did not identify arbitrary strangers worldwide. The important lesson is narrower: removing a visible name does not necessarily make detailed movement data anonymous.²⁶

Eye tracking is useful—not mind reading

Gaze can help control an interface or support task-specific research, but its meaning depends on context. Pupil responses, for example, depend on lighting as well as emotional arousal. A measurement is not a unique label for an emotion, let alone direct access to someone's beliefs or intentions.²⁷

Not every device collects every type of data. Tracking required to render an experience can still be sensitive; the WebXR specification explicitly discusses privacy, permissions and ways to escape immersive content. It does not establish that every installed XR application implements the same safeguards.²⁸

Bystanders matter too. Camera-equipped smart glasses are not necessarily AR displays, but they illustrate how recording can become less visible to people nearby. Make recording apparent and consider other people's permission before capturing or sharing classroom, clinical or private-space content.²⁹

In the EU, identifiable XR data may fall under data-protection law. Biometric data processed to uniquely identify a person receives special-category protection; health and other sensitive information may qualify separately. Not every raw gaze or movement reading automatically belongs to that category. The processing and purpose matter, alongside data minimization and protection by design.³⁰

Before buying a fleet of headsets, identify one task, one suitable application and one outcome worth improving.

For a school, the question might be whether a spatial lesson improves explanations of an unfamiliar diagram. For a clinic, it might be whether a defined program improves a relevant functional outcome and is acceptable to patients. “People liked it” is useful feedback, but it is not the whole evaluation.

1. Define the purpose and alternativeExplain why XR is being considered and what would otherwise be used. Identify who the activity is for and who may need a different route.
2. Inspect the exact applicationReview content accuracy, controls, data practices and supporting evidence. Record the version being evaluated; a familiar device name is not enough.
3. Run a supported trialBrief participants, check fit and space, offer a clear exit and document discomfort or access barriers as well as successes.
4. Review before expandingCompare meaningful outcomes with the alternative. Count setup, supervision, maintenance and support time. Change or stop the rollout if the benefit does not justify the burden.

Budget for the work around the device

Licenses, account management, charging, cleaning, storage, replacement parts, technical support and staff preparation all affect the real cost. Some applications can run offline; others depend on network services. Check the specific system rather than assuming broadband is always required—or that a downloaded app remains usable indefinitely.

Shared equipment also needs a cleaning process compatible with the manufacturer's instructions. Decide who resets accounts, checks accessories and makes sure the next participant is not shown the previous person's information. These operational details can determine whether an otherwise good lesson or treatment is practical.

Make a change log part of the plan

A software update can alter interaction, content, permissions or the data collected. Decide who reviews substantial changes before the application returns to use. For clinical systems, follow the applicable instructions and governance process; evidence for one program should not silently become endorsement of a different one.

Lighter equipment, clearer displays and more flexible controls are worthwhile goals. So are experiences that need fewer sensors, make fewer assumptions about bodies and remain useful outside a single vendor's ecosystem. A less spectacular system that more people can comfortably use may be the more important advance.

Research priorities include durable learning, transfer to real settings, comparison with good alternatives, adverse-effect reporting and repeated-use outcomes across diverse populations. In education and children's safety research, short studies and incomplete follow-up already limit what can be concluded.^{4, 22}

Open, accessible interaction and trustworthy data handling deserve the same attention as visual fidelity. Standards work such as WebXR addresses parts of this infrastructure, but a specification alone does not deliver a safe, inclusive experience.²⁸

Educational information, not individualized clinical or legal advice. Medical XR should be considered in the context of appropriate professional care and device instructions. Seek qualified advice about health conditions, persistent symptoms or clinical deployment, and review applicable privacy and safeguarding requirements for the setting.

Evidence & context

Sources and further reading

These sources distinguish specific trials, research syntheses, theoretical models and official guidance. A positive result applies to the intervention and outcome studied—not every XR product. External links open in a new tab.

1. US Food and Drug Administration. Augmented Reality and Virtual Reality in Medical Devices. Definitions, uses and device-specific benefits and risks.
2. Makransky G, Petersen GB. The Cognitive Affective Model of Immersive Learning (CAMIL). Educational Psychology Review (2021). Research-informed theoretical model.
3. Wu B, Yu X, Gu X. Effectiveness of immersive virtual reality using head-mounted displays on learning performance: A meta-analysis. British Journal of Educational Technology (2020).
4. Hamilton D et al. Immersive virtual reality as a pedagogical tool in education. Journal of Computers in Education (2021). Review of learning outcomes and experimental design.
5. Makransky G, Terkildsen TS, Mayer RE. Adding immersive virtual reality to a science lab simulation causes more presence but less learning. Learning and Instruction (2019).
6. Makransky G, Mayer RE. Benefits of Taking a Virtual Field Trip in Immersive Virtual Reality. Educational Psychology Review (2022). Greenland field trip within a climate curriculum.
7. Zingg L et al. Using Hologram-Based Augmented Reality in Anatomy Learning: The TEACHANATOMY Randomized Trial. Academic Medicine (2025).
8. Meyer OA, Omdahl MK, Makransky G. Investigating the effect of pre-training when learning through immersive virtual reality and video. Computers & Education (2019).
9. Parong J, Mayer RE. Learning science in immersive virtual reality. Journal of Educational Psychology (2018). Segmentation and summarizing experiment.
10. Freeman D et al. Automated psychological therapy using immersive virtual reality for treatment of fear of heights. The Lancet Psychiatry (2018). Randomized trial.
11. Freeman D et al. Automated virtual reality therapy to treat agoraphobic avoidance and distress in patients with psychosis (gameChange). The Lancet Psychiatry (2022).
12. US Food and Drug Administration. AR and VR medical devices: questions to consider. Patient and provider decision guidance.
13. US Food and Drug Administration. EaseVRx De Novo authorization (2021; PDF). Specific prescription indication and identified risks.
14. Garcia LM et al. Eight-week home behavioral-skills VR program for chronic low-back pain: randomized placebo-controlled trial. Journal of Medical Internet Research (2021).
15. Laver KE et al. Virtual reality for stroke rehabilitation. Cochrane Database of Systematic Reviews (2025 update; DOI ends pub5). Mostly nonimmersive interventions.
16. Li C et al. Cellphone augmented-reality training for upper-limb function and cognition after stroke. JMIR Serious Games (2021). Small, short-duration randomized trial.
17. Biswas N, Mukherjee A, Bhattacharya S. “Are you feeling sick?” A systematic literature review of cybersickness in virtual reality. ACM Computing Surveys (2024).
18. Hoffman DM et al. Vergence–accommodation conflicts hinder visual performance and cause visual fatigue. Journal of Vision (2008).
19. American Academy of Ophthalmology. Are Virtual Reality Headsets Safe for Eyes? (2024). Eye strain and visual-health context.
20. Apple. Safely use your Apple Vision Pro. Manufacturer guidance; consult the instructions for the device actually used.
21. Apple. If you experience motion sickness while using Apple Vision Pro. Symptoms, stopping and recovery guidance.
22. Bexson C, Oldham G, Wray J. Safety of virtual reality use in children: a systematic review. European Journal of Pediatrics (2024).
23. American Academy of Pediatrics. Virtual Reality and Children (2026). Family guidance and evidence gaps.
24. World Wide Web Consortium. XR Accessibility User Requirements (2021). Working Group Note describing diverse access needs.
25. eSafety Commissioner. Immersive technologies (2025 position paper and guidance). Embodied interaction, user controls and safety by design.
26. Nair V et al. Unique Identification of 50,000+ Virtual Reality Users from Head & Hand Motion Data. USENIX Security (2023).
27. Pan J et al. The effects of emotional arousal on pupil size depend on luminance. Scientific Reports (2024).
28. World Wide Web Consortium. WebXR Device API (Candidate Recommendation Draft, 9 June 2026). Privacy, permissions and immersive-session controls.
29. eSafety Commissioner. Smart glasses and online safety: When the camera disappears (2026). Recording and bystander privacy.
30. European Union. General Data Protection Regulation (2016), Articles 4, 5, 9, 25 and 32. Data categories, minimization and safeguards.