A GUIDE FOR TEACHERS AND THE PEOPLE WHO EQUIP THEM

Using immersive media

When a headset earns its cost, what has to happen in the twenty minutes after it comes off and the cheaper thing to try before committing to any of it.

A headset is the most striking thing most students will try on at school. That is the problem. A memorable hour and a durable lesson feel the same on the day, and only one of them is still there a month later.

This guide is about getting the second. Five sections hold thirty-one numbered items. Each states a claim, what to do about it and what to watch for. Each also carries a mark for how much evidence stands behind it.

WHO THIS IS FOR

For the teacher planning the lesson and the person deciding what to buy

Teachers, heads of department and the people who look after devices in a school. No technical background is needed. A lesson in mind is.

I wrote this for two readers who are usually in different meetings. One plans a lesson and wants to know whether a headset will help it. The other approves the spend and wants to know whether the lesson will happen at all. They need the same answers.

The guide works from the position I set out on the Immersive page. Presence, immersion and the research model behind them are defined in full in the primer on embodied, spatial and immersive learning, and only briefly here.

The longform version carries four appendices: definitions, the four cases at a glance, working templates and notes on the evidence.

THE COMMON BELIEF

“They remember the headset lesson, so it worked.”

They remember the headset. Whether they can explain the idea is a separate question, and the answer is often no.

In one study, university students who learned biology in a headset rated the lesson more interesting than students who saw the same content as slides. The slide group did better on the test (Parong & Mayer, 2018).

In another, moving a science simulation from a desktop screen into a headset raised the students’ sense of being there and lowered what they learned (Makransky et al., 2019). Feeling present and learning came apart.

Both studies are small and short. What they show is that enjoyment, and a vivid memory of the experience, are poor evidence of learning. The rest of this guide is built on that distinction.

THE SHAPE OF A HEADSET LESSON

The headset sits in the middle of the lesson

In a fifty-minute lesson, fifteen minutes inside is plenty. The twenty after decide what the fifteen were worth.

FIG 01: ONE WAY TO SPEND A FIFTY-MINUTE LESSON

The time around the headset is longer than the time inside it, and that is deliberate. Set-up takes five minutes even when it goes well. The ten minutes before give students a question to carry in. The twenty after turn what they saw into something they can say, write and use.

The minutes are illustrative. The proportions are the point. If a plan has the headset filling most of the period, something on either side of it has been cut, and it is usually the part that does the teaching.

WHAT THIS GUIDE STANDS ON

Most of the evidence is short and recent

Much of it measures one session, soon after it ends. Read every mark with that in mind.

FIG 02: THE MARKS ACROSS THIRTY-ONE ITEMS

The research on immersive media in education is large and young. A review that screened 1,220 articles and kept 97 named four barriers to wider use: cost, health and safety, fit with the curriculum and teacher training (Samala et al., 2025).

Novelty should be strongest on first contact. In a meta-analysis of VR in primary classes, interventions under two hours showed larger effects than longer ones (Villena-Taranilla et al., 2022). A novelty effect would produce that pattern, though the study does not test for one. So I treat any single-session gain with care, including the ones cited here.

HOW TO READ THE MARK ON EACH ITEM

Four marks

• REPLICATED Several independent studies agree and the direction of the effect holds
• SUPPORTED A real study or review stands behind it, narrower than the advice drawn from it
• CONVENTION Sound practice with no direct test, recommended on judgement
• CONTESTED The evidence is mixed, or points away from how the idea is usually sold

REPLICATEDThree of thirty-one Reviews and meta-analyses point the same way. Even here most studies are short, so the mark speaks to the direction of an effect more than its size.

SUPPORTEDSeventeen of thirty-one A study or a review exists and says something real, but less than the advice usually asks of it. Many rest on adults, on one session or on a single subject.

CONVENTIONNine of thirty-one Classroom practice that follows from the evidence without being tested itself: jobs for students without a headset, borrowing before buying, fixing a decision date. I recommend them, and say so.

CONTESTEDTwo of thirty-one Perspective-taking and the safety of headsets for children. In both the evidence is thin or mixed, and the items say what is known in its place.

The mark sits in the eyebrow of each numbered item, after the section letter and number. Where it is anything other than replicated, the panel says why.

THE CLAIM

Immersion earns its cost only where being there is the lesson

Everywhere else, something cheaper does the same job and leaves the student more of the thinking.

This is the position the whole guide argues from. Presence is a means. It counts when being somewhere lets a student do something they could not do any other way, and only then.

Section A

When a headset earns its cost

Four cases where being there is the lesson, and three reasons that sound like cases and are not.

• A When a headset earns its cost
• B The cheaper thing to try first
• C Inside the headset
• D The twenty minutes after
• E Judging whether it worked

Section A sorts the reasons for using a headset into the four that hold up and the three that only sound as if they do.

01

THE FOUR CASES

Section A

01

The four cases

Scale, danger, invisibility and perspective. A topic in one of these has a real claim on a headset.

I start every conversation about headsets with one question. What will the student do in there that they could not do anywhere else?

Four kinds of answer hold up. The thing is too big, too small, too fast or too slow to see. Doing it for real is dangerous. The thing has no everyday appearance. Or the point is to stand where someone else stands. Outside those four, a diagram, a model or a demonstration usually does the job for less and leaves more of the thinking to the student.

A · 01 · SUPPORTED

Name the action first

Before choosing a medium, write down what the student will do. A verb that needs being there can earn a headset. A verb that needs only seeing cannot.

FIG 03: TWO KINDS OF VERB IN AN OBJECTIVE

A headset offers two things a screen does not: the feeling of being somewhere and the sense of acting there. A leading model of learning in VR treats both as routes to learning, and warns that presence can raise interest and mental load together (Makransky & Petersen, 2021). The feeling does its work only through what the student does with it.

Write the objective as a verb the student performs: inspect, rehearse, decide, take a viewpoint. Then ask whether that verb needs a body in a place. A review of immersive VR studies found the medium suited learning built on handling and making things (Conrad et al., 2024).

Objectives written as “experience” or “explore”. They describe the medium, so any title will seem to meet them.

SUPPORTED · Makransky & Petersen, 2021 · Conrad et al., 2024

A · 02 · SUPPORTED

Scale

Some things are too large, too small, too fast or too slow to perceive. Standing beside them can show relations a page flattens.

FIG 04: THE BODY SEES ONLY A NARROW BAND

Spatial relations are where 3D tools have their clearest record. A meta-analysis of 36 studies in anatomy teaching found 3D visualisation beat other methods, more on spatial knowledge (d = 0.50) than on facts (d = 0.30) (Yammine & Violato, 2015). Those were 3D tools of several kinds, so the finding is about the third dimension, whatever the display. Headset studies show their larger effects in science and with school-age learners (Wu et al., 2020).

Use scale when the relation is the lesson: how a molecule fits a receptor, how a valley shapes a flood. Ask students to estimate a size or a distance before they go in, then check it inside.

Scale as spectacle. If nobody has to judge a distance or compare two sizes, a scale drawing on the hall floor teaches as much.

SUPPORTED · Yammine & Violato, 2015 · Wu et al., 2020

A · 03 · SUPPORTED

Danger

Where practising in the real setting would put someone at risk, a simulation lets students fail safely and try again.

FIG 05: ONE REAL ATTEMPT, OR SAFE ONES FIRST

A meta-analysis of simulation-based learning in higher education found a large average effect (g = 0.85), and larger gains when learners got support: examples for novices, reflection for those who knew more (Chernikova et al., 2020). Only 20 of its 145 studies used virtual reality, and simulations with no technology at all still worked well (g = 0.74). Much of the value lies in rehearsal with a reset button.

Social risk counts too. In a trial with 115 boys treated for aggressive behaviour, therapy that rehearsed provoking situations in VR beat the same therapy using role-play on four of seven outcomes (Alsem et al., 2023).

Use it where a mistake in the real setting hurts someone: a chemical spill, a fire drill, a busy road. Let students fail, reset and try again, then do the real version with an adult.

Rehearsal that never meets the real thing. A safe simulation is a step towards practice and cannot stand in for it.

SUPPORTED · Chernikova et al., 2020 · Alsem et al., 2023

A · 04 · SUPPORTED

Invisibility

Fields, flows and forces have no everyday appearance. Letting students see and handle them can make an abstraction concrete.

FIG 06: A FIELD THE STUDENT DRAWS BY HAND

Johnson-Glenberg (2018) proposes that the gain comes when a movement maps onto the idea. In a study she reports, 166 university students learning about electric fields used their arms to draw the arrows that show how a charge moves, on a large screen that tracked them. They did better than students who read or watched: slightly on a typed test and significantly on a test that asked them to draw. Nobody wore a headset.

Pick ideas whose shape matters: a magnetic field, air over a wing, current in a circuit. Have students act on the invisible thing, and predict what it will do before it is shown.

A beautiful rendering that students only look at. If the field moves while the student watches, a good animation on a screen does the same.

SUPPORTED · Johnson-Glenberg, 2018

A · 05 · CONTESTED

Perspective

Where standing in someone else’s position is the objective, VR can move feelings. Whether it changes understanding is far less clear.

FIG 07: WHAT VR MOVED ACROSS 43 STUDIES

A meta-analysis of 43 studies with 5,644 participants found VR raised emotional empathy but not cognitive empathy, the ability to understand another person’s view. It did no better than asking people to read about someone and imagine their situation. More immersive headsets did no better than simple cardboard viewers (Martingano et al., 2021). Perspective stays among the four cases because the viewpoint can be the objective itself. The mark is contested because little evidence shows VR changing understanding as well as feeling.

Use perspective when the viewpoint itself is what students must learn, and pair it with a task that asks them to reason from that position, in words, afterwards.

A strong feeling taken for a changed mind. Most studies measure straight after one session, and whose body the student occupies changes the result (Li & Kim, 2021). Section D returns to this.

CONTESTED · Martingano et al., 2021 · Li & Kim, 2021

02

WHAT DOES NOT COUNT

Section A

02

What does not count

Three reasons that sound like cases. Each is real, and none is enough on its own.

Most proposals for headsets rest on one of three reasons. The place is far away. The students will be excited. The money is there to spend.

Each sounds like a case. None of them names something the student will do, which is the test each of the four cases passed.

A · 06 · CONVENTION

Distance alone is no case

A place students cannot visit is a good reason to show them the place. It is rarely a reason for a headset.

A rule of thumb often quoted from VR research says to use it where the real experience would be dangerous, impossible, counterproductive or expensive (Bailenson, 2016, as cited in Johnson-Glenberg, 2018). Danger and impossibility map onto the four cases. Expense is weaker. A far-off place can usually be shown for much less as a 360 video on a screen, or as photographs and a map.

Ask what the student will do at the destination. If the answer is look around, use the screen. If it is judge a scale or take a viewpoint, the trip may belong in one of the four cases after all.

Virtual field trips justified by awe. Awe is pleasant and worth having, and on its own it fades.

CONVENTION · Johnson-Glenberg, 2018

A · 07 · SUPPORTED

Excitement wears off

The first time in a headset is thrilling. Results measured then will flatter the medium, and nobody yet knows by how much.

FIG 08: SHORTER INTERVENTIONS, LARGER EFFECTS

A meta-analysis of 21 studies of VR with primary-age children found an average effect of 0.64, and 1.11 for fully immersive VR. Interventions under two hours averaged 0.72, and longer ones 0.49 (Villena-Taranilla et al., 2022). Novelty would produce that pattern, though other explanations fit it too.

Plan for the tenth use rather than the first. Build the task so it would still be worth doing on an ordinary Tuesday in the second term, with a class that has seen the headsets before.

Enthusiasm reported as impact. How much of the measured advantage is novelty, and how fast it decays with repeated use, is one of the open questions I keep on the Immersive page.

SUPPORTED · Villena-Taranilla et al., 2022

A · 08 · SUPPORTED

Count what the headset displaces

The price of the headsets is the smallest cost. Teacher time, licences, upkeep, supervision and lost lesson time are larger.

FIG 09: THE PRICE TAG IS THE PART YOU CAN SEE

A review of 97 studies of VR in education named four barriers to wider use: cost, health and safety, fit with the curriculum and teacher training (Samala et al., 2025). Only the first appears on an invoice, and even that one keeps arriving as licences and replacements.

Before any purchase, list what the lesson time would otherwise be spent on and who will charge, update, clean and store the devices. Put a name against each job.

Equipment bought because a budget had to be spent by a date. Headsets in a cupboard cost the same as headsets in use.

SUPPORTED · Samala et al., 2025

Section B

The cheaper thing to try first

A ladder of media, and the habit of climbing it only as far as the learning needs.

• A When a headset earns its cost
• B The cheaper thing to try first
• C Inside the headset
• D The twenty minutes after
• E Judging whether it worked

Section B sets out the media below the headset, and why the habit of trying them first is what makes the case for a headset believable.

03

THE LADDER

Section B

03

The ladder

Climb from the cheapest medium that could carry the lesson, and stop at the first one that does.

I ask what learning action the experience is meant to enable. Then I ask what the cheapest medium is that would enable the same action. If the answer is a diagram, we make the diagram.

Choosing the smaller option when it does the same job is what makes the case believable on the days when the headset really is the right call.

B · 01 · REPLICATED

Climb the ladder one step at a time

Diagram, video, 3D model, desktop simulation, headset. Each step costs more. Climb only as far as the learning action needs.

FIG 10: FIVE STEPS, EACH DEARER THAN THE LAST

The argument about media is old. Clark (1994) held that media never influence learning, only the methods they carry. Kozma (1994) replied that some methods need particular media. Both survive here: choose the method first, then the cheapest medium able to carry it.

Meta-analyses find headsets help on average, by a small margin. Across 35 studies the effect against less immersive approaches was 0.24 (Wu et al., 2020). Across 48 studies it was again small, at g = 0.38 (Coban et al., 2022).

Start at the bottom. Try the diagram or the model, and move up only when you can name what the next step adds.

Skipping to the top because the headsets are already in the building.

REPLICATED · Wu et al., 2020 · Coban et al., 2022

B · 02 · CONVENTION

Use the tablets you already own

Augmented reality on phones and tablets puts a 3D model on a desk for a fraction of the cost and keeps the room in view.

Augmented reality lays a digital object over the real room, seen through a camera. Students can walk round a model of the heart on a desk, or turn a cross-section of a volcano on a table. Because they can still see each other, they can point, argue and share one screen. That makes it a natural first step up from a diagram.

Check which devices in school can run it and try one title with one class. Give each pair one device and two jobs: one moves the model, the other says what changes.

AR as a sticker. A model that floats on the desk and does nothing teaches no more than the picture in the textbook.

CONVENTION · Classroom practice, untested as a comparison

B · 03 · SUPPORTED

Run the simulation on a screen first

A desktop simulation gives most of what a headset gives for learning and makes it easy to ask for a prediction before every run.

FIG 11: PREDICT, TEST, EXPLAIN, APPLY

Discovery learning with simulations works when learners get support. Left alone, they struggle to form hypotheses and design fair tests (de Jong & van Joolingen, 1998). And when one science simulation moved from a desktop into a headset, students felt more present and learned less (Makransky et al., 2019).

Ask for a written prediction before each run and change one variable at a time. End with a problem the simulation never showed.

Free play with sliders. Students will move everything at once and learn mainly that the simulation responds.

SUPPORTED · de Jong & van Joolingen, 1998 · Makransky et al., 2019

B · 04 · CONVENTION

Borrow before you buy

Try one unit with a few borrowed headsets for one term, and fix the date you will decide before you start.

FIG 12: ONE UNIT, ONE TERM, ONE DECISION

A trial is a test that can fail. If stopping was never a real option, it was a launch with a smaller budget. The point of borrowing is that saying no costs nothing.

Choose one unit where the topic sits in one of the four cases. Set the question and the comparison before the headsets arrive. Log every session. Decide on the date you fixed at the start.

A trial judged by how much students liked it. Section E sets out what to look at instead.

CONVENTION · Practice guidance, untested as a comparison

Section C

Inside the headset

Choosing what students will see, and using the minutes they spend seeing it.

• A When a headset earns its cost
• B The cheaper thing to try first
• C Inside the headset
• D The twenty minutes after
• E Judging whether it worked

Section C covers the two things a teacher controls once a headset is justified: which title students use, and how the minutes inside are spent.

04

CHOOSING CONTENT

Section C

04

Choosing content

Most of what a headset lesson can be is fixed on the day the title is chosen.

Teachers rarely build immersive content. They choose it, usually from a catalogue written to sell. That makes the choice the biggest design decision in the lesson, and the one most often made from a trailer.

C · 01 · CONVENTION

Play it through as a student would

Try the whole title on the device students will use, at their pace, before planning anything around it.

Trailers show the best ninety seconds. A student meets the menus, the loading screens, the tutorial and the moment the controller stops responding.

Time how long it takes to reach the part that matters. Note every word students must read, every controller action they must learn and every place someone could get stuck.

Content that teaches something slightly different from the syllabus. A title on the heart may label parts your students are never taught, or skip the ones they are.

CONVENTION · Practice guidance, untested as a comparison

C · 02 · SUPPORTED

Check what the hands do

A title where students act on the idea with their bodies is worth more than one where they press a button to watch it happen.

FIG 13: THE SAME GEAR, TWO KINDS OF MOVEMENT

Johnson-Glenberg (2018) separates low embodiment, pressing a key to spin a gear, from high embodiment, turning your arm so the gear follows your speed and direction. Students who physically felt the forces on a spinning wheel later scored higher on a physics quiz than students who only watched (Kontra et al., 2015).

Ask whether each movement in the title matches the concept: turning a gear by turning, stretching a spring by pulling. A controller that only points and clicks is a mouse with a strap.

Movement for its own sake. Skulmowski and Rey (2018) warn against expecting more learning from more movement. The movement has to belong to the task.

SUPPORTED · Johnson-Glenberg, 2018 · Kontra et al., 2015

C · 03 · SUPPORTED

Read the terms before the content

Headsets collect more than a laptop does. Check what accounts students need and what data leaves the room.

A headset records how its wearer moves, and movement works like a fingerprint. In one study, a model trained on head and hand motion picked out one person from more than 55,000 users with 94 per cent accuracy from 100 seconds of data, and 73 per cent from 10 seconds (Nair et al., 2023).

Before a trial, find out whether each student needs a personal account, what age the maker sets for one, what the device records and where it goes. Ask whoever handles data protection in your school to read the terms with you.

Licences that run per device or per year. A cheap title can become a recurring cost, and a lapsed licence can strand a unit halfway through a term.

SUPPORTED · Nair et al., 2023

05

THE TIME INSIDE

Section C

05

The time inside

Short, tasked and watched. Most of what goes wrong in a headset lesson goes wrong here.

Once the headsets are on, the teacher can no longer see what students see, and students cannot see the teacher. Everything that matters has to be arranged before that moment: the task, the jobs, the signal to stop.

C · 04 · SUPPORTED

Give a task, and leave something to build

An open world to explore produces little. A task with something to find, decide or explain gives the experience a point.

FIG 14: A TOUR WANDERS, A TASK HAS TARGETS

Rich environments fill working memory quickly (Sweller, 2011). Interesting but irrelevant details tend to reduce learning (Rey, 2012). A headset supplies both in quantity. If it also supplies the answer, the student has nothing left to construct, and construction is where the learning happens.

Give a task card with two or three targets: find, measure, decide, say why. Leave the conclusion for students to draw after the headset comes off.

“Explore and see what you notice.” Students will notice the most vivid thing in the scene, which is rarely the thing you meant.

SUPPORTED · Sweller, 2011 · Rey, 2012

C · 05 · SUPPORTED

Cut the time inside into short segments

Stopping at intervals for students to write about what they just saw improves learning, and costs nothing in interest.

FIG 15: THE SAME FIFTEEN MINUTES, TWO WAYS

In the paper where slides beat a VR biology lesson, a second experiment split the VR lesson into segments with a written summary after each. Those students did significantly better than students who watched it straight through, with no loss of interest or motivation (Parong & Mayer, 2018).

Find two or three natural pauses in the title. At each, students lift the headset and write a sentence or two on what they saw and why it matters.

Long unbroken sessions. They add discomfort as well as load, and time in the headset is one of the factors linked to sickness (Saredakis et al., 2020).

SUPPORTED · Parong & Mayer, 2018 · Saredakis et al., 2020

C · 06 · CONVENTION

Give everyone without a headset a job

Most classes have fewer headsets than students. The students waiting are the ones most likely to drift.

FIG 16: FOUR JOBS AROUND ONE HEADSET

With six headsets and a class of thirty, four in five students are outside a headset at any moment. Without a job, the lesson is happening to one student in five. The class is invented. The arithmetic holds wherever there is one headset for every five students.

Group students around each headset and rotate four jobs. One wears it and does the task. One guards the space around them. One writes down what the wearer says. One asks the questions on the task card. A review of group work in VR describes the same shape: a shared goal the group must work on together (Paulsen et al., 2024).

The queue. A student watching a classmate’s screen with nothing to do is waiting, however attentive they look.

CONVENTION · Paulsen et al., 2024

C · 07 · REPLICATED

Stop at the first sign of sickness

Some students will feel unwell. The answer is to stop early, sit down and never push through.

FIG 17: MOTION SEEN BUT NOT FELT

A review of 55 studies with 3,016 participants found sickness scores high against recommended cut-offs, linked to what users saw, how they moved and how long they stayed in (Saredakis et al., 2020). Current headsets cause less of it than older ones, but some symptoms remain as intense. The leading explanation is a mismatch between motion seen and motion felt (Caserman et al., 2021).

Agree one stop signal before anyone starts. Prefer seated use and titles that move by teleport or snap turns. Keep a plain-screen version of the task, so a student who stops still learns.

Pressure to finish. A student who feels ill and keeps going so as not to miss out is the case this rule exists for.

REPLICATED · Saredakis et al., 2020 · Caserman et al., 2021

C · 08 · CONTESTED

Treat age limits as a floor, and supervise

The evidence on children and headsets is thin. Short, supervised use has shown few harms. Nobody knows about long, repeated use.

A review of 26 studies of children under 14 found mostly mild effects, such as cybersickness. Some children with a lazy eye had double vision that cleared afterwards. Only two studies used a validated safety measure, and the effects of repeated use are unknown (Bexson et al., 2024). Fears of seizures from 3D displays have little support (Tychsen & Thio, 2020).

Follow the maker’s minimum age, keep sessions short, have an adult present and ask families about known conditions. Make sure every task can be done another way, so a student who opts out loses nothing.

A lack of evidence read as evidence of safety. The studies are few and short, which is why this item is marked contested.

CONTESTED · Bexson et al., 2024 · Tychsen & Thio, 2020

Section D

The twenty minutes after

Where the experience becomes something a student can explain, use and keep.

• A When a headset earns its cost
• B The cheaper thing to try first
• C Inside the headset
• D The twenty minutes after
• E Judging whether it worked

Section D is the part of the lesson most often cut. It is also where most of the teaching happens.

06

THE DEBRIEF

Section D

06

The debrief

The conversation after the headset comes off does much of the teaching. It is also the first thing cut when time runs short.

I plan the conversation afterwards before I plan the experience. If there is no time for the debrief, there is no time for the headset either.

The six items here turn what each student saw into something the class can say, write, compare and use on a problem the headset never showed.

D · 01 · SUPPORTED

Plan the debrief before the experience

Write the questions for the twenty minutes after first. Then choose the experience that gives students something to answer them with.

FIG 18: THE TWENTY MINUTES AFTER, ONE WAY

The studies point one way. Written summaries between segments improved learning from VR (Parong & Mayer, 2018). Reflection phases raised the effect of simulations for learners with some prior knowledge (Chernikova et al., 2020). Fiorella and Mayer (2016) review eight generative activities, from summarising to self-explaining, each with experimental support.

Draft the questions first. What did you notice? What surprised you? What does it mean? Then check that the experience gives every student material to answer them.

A debrief scheduled for the last five minutes. It will be cut, and the lesson will end on the most exciting part and the least understood.

SUPPORTED · Parong & Mayer, 2018 · Chernikova et al., 2020

D · 02 · REPLICATED

Write or draw before anyone talks

Each student puts down what they saw before the discussion starts, so the first confident voice does not become the class’s memory.

Summarising, drawing and explaining to yourself each have a body of experimental support as ways to make learners select and organise what they met (Fiorella & Mayer, 2016). Writing first also gives every student something of their own to bring to the conversation.

Give three minutes of silence and a prompt. Draw what you saw. Label the part that changed. Write one sentence on why it changed.

A discussion that opens with “So, what did you all think?” The most fluent student answers, and the rest agree.

REPLICATED · Fiorella & Mayer, 2016

D · 03 · CONVENTION

Turn private views into a shared account

Each student saw a slightly different world. Pool the views, and the differences become material for the lesson.

FIG 19: FOUR PRIVATE VIEWS, ONE SHARED ACCOUNT

A headset is a private experience. Two students in the same title look at different things in a different order. A review of group work in VR found current designs weak at supporting social interaction (Paulsen et al., 2024), and a review of VR laboratories found them still short of real labs on collaboration (Dodevska et al., 2025).

Give each student or group something different to look for. Then build one account together, on paper or the board, in which every view is needed.

Students reporting in turn without listening. A shared account needs a question that no single report can answer.

CONVENTION · Paulsen et al., 2024 · Dodevska et al., 2025

D · 04 · SUPPORTED

Name the general idea out loud

An experience teaches one case. Students need help to see what the case is an example of, and where else it holds.

FIG 20: ONE CASE, THE IDEA, A NEW CASE

Transfer rarely happens by itself. Perkins and Salomon (1988) call the deliberate version bridging: helping learners name the general principle and look for other places it applies. Varied worked examples help learners carry a method to new problems (Paas & van Merriënboer, 1994). Memory also works best when the test asks for the kind of processing the learning did (Morris et al., 1977), which is a reason to practise the idea in the form students will need it.

Ask two questions in order. What is this an example of? What else is an example of it? Then give a second case, in a different form, that the idea explains.

Trusting the experience to carry the idea by itself. Vivid cases are remembered as cases.

SUPPORTED · Perkins & Salomon, 1988 · Paas & van Merriënboer, 1994

D · 05 · SUPPORTED

Finish on a problem the headset never showed

End with a task in a different form: a diagram to complete, a case to explain, a prediction to make. That is where you see what transferred.

In one study, 141 retail trainees had four 90-minute VR sessions built as a bridge between classroom theory and the workplace. They did better than trainees taught in the usual way, on knowledge and on confidence to apply it at work (Kablitz, 2025). The design was quasi-experimental, so the result is encouraging and unsettled.

Keep the last seven minutes for a short problem on paper that uses the idea in a new setting. Collect it. It is the best evidence you will get that day.

Ending on the headset. If the last thing students do is take it off, the last thing they remember is the experience.

SUPPORTED · Kablitz, 2025

D · 06 · SUPPORTED

After a perspective lesson, ask what the person needed

Move students from how it felt to what the other person needed. That is the step from feeling to understanding.

FIG 21: WHOSE BODY IT IS CHANGES WHAT FOLLOWS

In a study of 128 people, experiencing organ failure in VR as themselves led to personal distress and other kinds of help, such as giving money or time. Experiencing it as another patient led to more empathy and more willingness to donate a kidney (Li & Kim, 2021). How the perspective is set up changes what it produces, and across studies VR moves feeling more reliably than understanding (Martingano et al., 2021).

Afterwards, ask students to describe the situation from the other person’s side in writing, then say what that person needed from others and why.

Ending on the emotion. Whether perspective-taking in VR changes how someone reasons months later, or leaves a strong feeling that fades, is still open.

SUPPORTED · Li & Kim, 2021 · Martingano et al., 2021

Section E

Judging whether it worked

How to tell learning from enthusiasm, and how to decide what happens next.

• A When a headset earns its cost
• B The cheaper thing to try first
• C Inside the headset
• D The twenty minutes after
• E Judging whether it worked

Section E is for whoever has to say, at the end of a trial, whether the headsets stay.

07

WHAT COUNTS AS EVIDENCE

Section E

07

What counts as evidence

Enjoyment is easy to measure and tells you little. Learning outside the headset is harder to measure and tells you what you need.

The question at the end of a trial is narrow. Did this help students learn this objective better than a workable alternative, at a cost the school can keep paying? Everything else is useful context and weak evidence.

E · 01 · SUPPORTED

Judge by what students can explain afterwards

Presence, enjoyment and engagement are real and pleasant. None of them is evidence of learning.

FIG 22: MORE PRESENT, LESS LEARNED

When one simulation moved from a desktop to a headset, 52 university students reported more presence (d = 1.30) and learned less (d = 0.80), with higher measured mental load (Makransky et al., 2019). In another study, students rated a VR lesson more interesting than slides, and the slides produced better test scores (Parong & Mayer, 2018).

Decide in advance what a student should be able to explain, draw or do after the lesson. Check it with every student, on paper, away from the headset.

Enjoyment surveys reported as outcomes. Ask about comfort and confusion, which are useful, and measure learning separately.

SUPPORTED · Makransky et al., 2019 · Parong & Mayer, 2018

E · 02 · SUPPORTED

Change only the medium

A fair comparison gives both classes the same task, the same guidance and the same time. Only the display differs.

FIG 23: A CONFOUNDED COMPARISON AND A FAIR ONE

A review of 38 studies comparing immersive VR with other media for teaching STEM subjects found only 26 per cent of the comparisons fully controlled, and 40 per cent with at least one confound, where something besides the medium differed between groups (Lawson et al., 2024). When the VR class also gets a better lesson, the headset takes the credit for the lesson.

To learn whether the headset helped, teach the same lesson to two classes with the same task card, the same debrief and the same time. Change only the display.

A new VR lesson compared with an old lecture. You will learn that active lessons beat passive ones, which was already known.

SUPPORTED · Lawson et al., 2024

E · 03 · CONVENTION

Look for learning outside the headset

A score inside the experience shows the student learned the experience. The test that matters happens somewhere else, later.

FIG 24: EVIDENCE OF LEARNING, WEAK TO STRONG

Stokes and Baer (1977) argued that generalisation has to be planned for and checked, because it seldom happens on its own. One trial gave 26 adults job-interview practice with a virtual interviewer on a computer. It measured the result away from the screen, in a live role-play interview afterwards, and the practice group improved more (Smith et al., 2014). That is the shape of measure to copy.

Use a task outside the headset, in a different form, ideally weeks later and marked without knowing which class the student was in.

Scores taken from inside the title. They show that students learned to play it.

CONVENTION · Stokes & Baer, 1977 · Smith et al., 2014

E · 04 · CONVENTION

Watch the room as well as the result

Record how often the headsets were used, what everyone else was doing and where the teacher’s time went.

A lesson can teach well and still be unworkable. If set-up eats ten minutes, the teacher spends the lesson fixing devices and most of the class waits, the headsets will not survive a second term, whatever the scores say.

Keep a short log for every session: date, class, minutes in headsets, problems. Sit in on one lesson and note what the students without a headset are doing, and what the teacher is doing. Ask students a few questions about comfort and clarity.

Use counted as success. A headset used every week for the wrong reason is a recurring cost with nothing to show for it.

CONVENTION · Practice guidance, untested as a comparison

E · 05 · CONVENTION

Decide on a date set in advance

Fix the date and the three possible answers before the trial starts: widen it, adjust it or stop.

FIG 25: THREE ANSWERS, CHOSEN BEFORE THE DATA

Once money and effort are spent, it is hard to judge a trial on its results alone. Setting the date and the criteria first keeps the review honest, and makes stopping an ordinary outcome.

Write down before the start what result would justify widening the trial, what would justify changing it and what would justify stopping. Bring the teachers and a few students to the review.

A review with only one possible answer. If stopping was never on the table, write that down too, and stop calling it a trial.

CONVENTION · Practice guidance, untested as a comparison

08

PUTTING IT TO WORK

08

Putting it to work

Seven questions to ask before the headsets come out of the box.

The guide ends where a decision starts. Ask these before anything is bought or booked, and ask them again after the first lesson.

BEFORE ANYTHING IS BOUGHT OR BOOKED

Seven questions to ask first

• 01 What will students do that they could not do any other way?
• 02 Is the topic a case of scale, danger, invisibility or perspective?
• 03 What is the cheapest medium that allows the same action?
• 04 What will students build in their heads, and what will the experience build for them?
• 05 Where are the twenty minutes after, and what are they for?
• 06 What does this displace from the timetable?
• 07 How will you know, weeks later, whether it worked?

The first six come from the position I hold on immersive media. The seventh is the one most often left until it is too late to answer.

THE REFERENCE LAYER

Four appendices hold the detail

Definitions, the four cases at a glance, working templates and notes on the evidence are in the longform version.

Appendix A defines the terms. Appendix B sets the four cases beside the reasons that do not count. Appendix C holds templates for a content review, a lesson plan, a debrief, a trial plan and a few questions for students. Appendix D records what the evidence in this guide can and cannot carry.

Name the action first.

Try the cheaper thing.

If there is no time for the twenty minutes after, there is no time for the headset.

KEEP READING

Where to go next

Where the position comes from, where the terms are defined and where the other guides live.

• POSITION geraldajam.com/ideas/immersive
• PRIMER Embodied, spatial and immersive learning
• GUIDES geraldajam.com/guides

For the research, start with Makransky and Petersen (2021) for a model of how VR affects learning, Lawson et al. (2024) on why so many comparisons mislead, and Martingano et al. (2021) on empathy. Fiorella and Mayer (2016) on generative activities applies to any medium, which is rather the point.

Reference layer

Appendix

Appendix A

Definitions

The terms this guide uses, in the sense it uses them. Presence, immersion and the model behind them are covered in full in the primer on embodied, spatial and immersive learning.

01360 video Video recorded in every direction, which a viewer can look around in on a screen or in a headset. It shows a place and offers little to act on.

02Agency The sense of controlling your own actions in an environment. With presence, one of the two routes by which immersive VR is thought to affect learning (Makransky & Petersen, 2021).

03Augmented reality (AR) Digital objects laid over a live view of the real room, usually through the camera of a phone or tablet.

04Bridging Deliberately helping learners name the general idea in a case and find other places it applies (Perkins & Salomon, 1988).

05CAMIL The Cognitive Affective Model of Immersive Learning, which links presence and agency to learning through six factors: interest, motivation, self-efficacy, embodiment, mental load and self-regulation (Makransky & Petersen, 2021).

06Confound Any difference between two groups other than the one being tested, such as a different activity for the VR class (Lawson et al., 2024).

07Cybersickness Nausea, headache, eye strain or dizziness during or after VR use, mostly put down to a mismatch between motion seen and motion felt (Caserman et al., 2021).

08Debrief The writing and conversation after an experience that turn it into something students can explain and use.

09Embodiment Here, learning in which a physical movement maps onto the concept being learned (Johnson-Glenberg, 2018).

10Generative activity A task that makes learners select, organise and connect what they met, such as summarising, drawing or explaining (Fiorella & Mayer, 2016).

11Immersion A property of the system: how fully its display and tracking surround the senses (Slater & Wilbur, 1997).

12Media comparison study A study that tests one medium against another for the same lesson.

13Mixed reality A view in which real and digital objects share the scene. It sits on a continuum that runs from the real room to fully virtual reality (Milgram & Kishino, 1994).

14Novelty effect A gain that comes from the newness of a medium and fades with repeated use.

15Presence The feeling of being in the virtual place. A response of the user, as distinct from immersion, which belongs to the system (Slater & Wilbur, 1997).

16Seductive details Interesting but irrelevant material, which tends to reduce learning (Rey, 2012).

17Six degrees of freedom Tracking of both where you look and where you move. Three degrees track only where you look, so walking forward does nothing.

18Transfer Using what was learned in a new setting, or on a new form of problem.

19Virtual reality (VR) A display, usually a headset, that replaces the view of the room with a computer-generated one.

Appendix B

The four cases at a glance

The four cases that can earn a headset, and the three reasons that cannot on their own, with the cheaper thing to try first.

Appendix C

Working templates

Short forms that make the habits in this guide routine. Copy them into whatever the school already uses.

01A content review The learning action, written as a verb. Which of the four cases it is, if any. Minutes from start-up to the part that matters. What the hands do, and whether it matches the concept. Text students must read and controller actions they must learn. Accounts, data collected and licence terms. Where the content differs from the syllabus.

02A lesson plan around a headset The question students carry in. The task card, with two or three targets. A job for everyone without a headset. The stop signal, and the plain-screen version of the task. The pauses for writing, and when they fall. The twenty minutes after, drafted first.

03A debrief Three minutes writing or drawing alone. Pairs compare and find one difference. The class names the general idea. “What is this an example of? What else is?” A new problem on paper, without the headset.

04A trial plan One unit and one objective, in one of the four cases. The comparison: the same lesson without the headset. A session log: date, class, minutes in headsets, problems. One observed lesson: what the others did, and what the teacher did. An outcome task, marked without knowing the class. The decision date, and what result would justify widening, adjusting or stopping.

05Questions for students How comfortable was the headset, from one to four? What were you looking for inside? Did you feel dizzy, sick or sore-eyed? When? What did you do while you waited? What do you understand now that you did not before?

Appendix D

Notes on the evidence

What this guide relies on, how strong it is and where to be careful.

01Who was studied Most studies cited here involve university students or adults, in a single session, measured soon after. Studies of school-age children are fewer (Villena-Taranilla et al., 2022, is the main one here), and studies of repeated use across a term are rare.

02Novelty No study cited here isolates a novelty effect. The pattern in Villena-Taranilla et al. (2022), larger effects for shorter interventions, fits one and has other explanations, such as shorter studies using narrower tests.

03Confounds Lawson et al. (2024) found most VR comparisons in STEM education were not fully controlled. Read the average effects in Wu et al. (2020) and Coban et al. (2022) with that in mind.

04Details taken at second hand The electric-field study under A·04 is described from the account in Johnson-Glenberg (2018). The original study was not read. The rule of thumb under A·06 is cited as she reports it too. The figures for Coban et al. (2022), 48 studies and an effect of g = 0.38, were checked against its abstract.

05Empathy Martingano et al. (2021) pool studies that mostly measure empathy straight after one session. Li and Kim (2021) is a single study. Whether any of the effects last is unknown. That, and the gap between feeling and understanding, is why A·05 is marked contested.

06Safety Bexson et al. (2024) found only two of 26 studies used a validated safety measure. The remedies for sickness under C·07 (seated use, teleport, snap turns) follow from the mismatch explanation in Caserman et al. (2021) and the link to movement in Saredakis et al. (2020). They are sensible practice. This guide cites no test of them as remedies.

07Practice items Items marked convention, such as jobs for students, borrowing before buying and a fixed decision date, are sound practice with no direct test. They are my recommendations, and a particular school may have good reason to do otherwise.

08What was left out Product names, prices, makers’ age rules and account terms change faster than a guide, so none appear here. Check them at the time of any purchase. Findings specific to students with special educational needs are left out of the panels, and Smith et al. (2014), a study of adults with autism who practised on a computer, is used only as an example of measuring learning away from the simulation. Research on general screen time is left out because it says little about headsets.

09Illustrative figures The fifty-minute lesson, the twenty-minute debrief, the class of thirty with six headsets and the ranking of evidence under E·03 are invented to show a pattern. They do not describe a real lesson, school or study.

Source trail

References

The works this guide draws on. Entries follow APA 7th edition, and links point to the DOI or source page. Each entry was checked against a publisher record or a DOI registry, and each claim against the source or its abstract, in November 2026. Where a detail comes from a secondary account, Appendix D says so.

1. Alsem, S. C., van Dijk, A., Verhulp, E. E., Dekkers, T. J., & De Castro, B. O. (2023). Treating children’s aggressive behavior problems using cognitive behavior therapy with virtual reality: A multicenter randomized controlled trial. Child Development, 94(6), e344–e361. https://doi.org/10.1111/cdev.13966
2. Bexson, C., Oldham, G., & Wray, J. (2024). Safety of virtual reality use in children: A systematic review. European Journal of Pediatrics, 183(5), 2071–2090. https://doi.org/10.1007/s00431-024-05488-5
3. Caserman, P., Garcia-Agundez, A., Gámez Zerban, A., & Göbel, S. (2021). Cybersickness in current-generation virtual reality head-mounted displays: Systematic review and outlook. Virtual Reality, 25(4), 1153–1170. https://doi.org/10.1007/s10055-021-00513-6
4. Chernikova, O., Heitzmann, N., Stadler, M., Holzberger, D., Seidel, T., & Fischer, F. (2020). Simulation-based learning in higher education: A meta-analysis. Review of Educational Research, 90(4), 499–541. https://doi.org/10.3102/0034654320933544
5. Clark, R. E. (1994). Media will never influence learning. Educational Technology Research and Development, 42(2), 21–29. https://doi.org/10.1007/BF02299088
6. Coban, M., Bolat, Y. I., & Goksu, I. (2022). The potential of immersive virtual reality to enhance learning: A meta-analysis. Educational Research Review, 36, 100452. https://doi.org/10.1016/j.edurev.2022.100452
7. Conrad, M., Kablitz, D., & Schumann, S. (2024). Learning effectiveness of immersive virtual reality in education and training: A systematic review of findings. Computers & Education: X Reality, 4, 100053. https://doi.org/10.1016/j.cexr.2024.100053
8. de Jong, T., & van Joolingen, W. R. (1998). Scientific discovery learning with computer simulations of conceptual domains. Review of Educational Research, 68(2), 179–201. https://doi.org/10.3102/00346543068002179
9. Dodevska, M., Zdravevski, E., Chorbev, I., Kostoska, M., Branco, F., Coelho, P. J., Pires, I. M., & Lameski, P. (2025). Virtual reality as a learning tool: Evaluating the use and effectiveness of simulation laboratories in educational settings. Social Sciences & Humanities Open, 12, 101742. https://doi.org/10.1016/j.ssaho.2025.101742
10. Fiorella, L., & Mayer, R. E. (2016). Eight ways to promote generative learning. Educational Psychology Review, 28(4), 717–741. https://doi.org/10.1007/s10648-015-9348-9
11. Johnson-Glenberg, M. C. (2018). Immersive VR and education: Embodied design principles that include gesture and hand controls. Frontiers in Robotics and AI, 5, 81. https://doi.org/10.3389/frobt.2018.00081
12. Kablitz, D. (2025). Bridging theory and practice with immersive virtual reality: A study on transfer facilitation in VET. Education Sciences, 15(8), 959. https://doi.org/10.3390/educsci15080959
13. Kontra, C., Lyons, D. J., Fischer, S. M., & Beilock, S. L. (2015). Physical experience enhances science learning. Psychological Science, 26(6), 737–749. https://doi.org/10.1177/0956797615569355
14. Kozma, R. B. (1994). Will media influence learning? Reframing the debate. Educational Technology Research and Development, 42(2), 7–19. https://doi.org/10.1007/BF02299087
15. Lawson, A. P., Martella, A. M., LaBonte, K., Delgado, C. Y., Zhao, F., Gluck, J. A., Munns, M. E., Wells LeRoy, A., & Mayer, R. E. (2024). Confounded or controlled? A systematic review of media comparison studies involving immersive virtual reality for STEM education. Educational Psychology Review, 36(3), 69. https://doi.org/10.1007/s10648-024-09908-8
16. Li, B. J., & Kim, H. K. (2021). Experiencing organ failure in virtual reality: Effects of self- versus other-embodied perspective taking on empathy and prosocial outcomes. New Media & Society, 23(8), 2144–2166. https://doi.org/10.1177/1461444821993122
17. Makransky, G., & Petersen, G. B. (2021). The Cognitive Affective Model of Immersive Learning (CAMIL): A theoretical research-based model of learning in immersive virtual reality. Educational Psychology Review, 33(3), 937–958. https://doi.org/10.1007/s10648-020-09586-2
18. Makransky, G., Terkildsen, T. S., & Mayer, R. E. (2019). Adding immersive virtual reality to a science lab simulation causes more presence but less learning. Learning and Instruction, 60, 225–236. https://doi.org/10.1016/j.learninstruc.2017.12.007
19. Martingano, A. J., Hererra, F., & Konrath, S. (2021). Virtual reality improves emotional but not cognitive empathy: A meta-analysis. Technology, Mind, and Behavior, 2(1), 7–21. https://doi.org/10.1037/tmb0000034
20. Milgram, P., & Kishino, F. (1994). A taxonomy of mixed reality visual displays. IEICE Transactions on Information and Systems, E77-D(12), 1321–1329. https://globals.ieice.org/en_transactions/information/10.1587/e77-d_12_1321/_p
21. Morris, C. D., Bransford, J. D., & Franks, J. J. (1977). Levels of processing versus transfer appropriate processing. Journal of Verbal Learning and Verbal Behavior, 16(5), 519–533. https://doi.org/10.1016/S0022-5371(77)80016-9
22. Nair, V., Guo, W., Mattern, J., Wang, R., O’Brien, J. F., Rosenberg, L., & Song, D. (2023). Unique identification of 50,000+ virtual reality users from head & hand motion data. In 32nd USENIX Security Symposium (pp. 895–910). USENIX Association. https://arxiv.org/abs/2302.08927
23. Paas, F. G. W. C., & van Merriënboer, J. J. G. (1994). Variability of worked examples and transfer of geometrical problem-solving skills: A cognitive-load approach. Journal of Educational Psychology, 86(1), 122–133. https://doi.org/10.1037/0022-0663.86.1.122
24. Parong, J., & Mayer, R. E. (2018). Learning science in immersive virtual reality. Journal of Educational Psychology, 110(6), 785–797. https://doi.org/10.1037/edu0000241
25. Paulsen, L., Dau, S., & Davidsen, J. (2024). Designing for collaborative learning in immersive virtual reality: A systematic literature review. Virtual Reality, 28(1), 63. https://doi.org/10.1007/s10055-024-00975-4
26. Perkins, D. N., & Salomon, G. (1988). Teaching for transfer. Educational Leadership, 46(1), 22–32.
27. Rey, G. D. (2012). A review of research and a meta-analysis of the seductive detail effect. Educational Research Review, 7(3), 216–237. https://doi.org/10.1016/j.edurev.2012.05.003
28. Samala, A. D., Rawas, S., Rahmadika, S., Criollo-C, S., Fikri, R., & Sandra, R. P. (2025). Virtual reality in education: Global trends, challenges, and impacts – game changer or passing trend? Discover Education, 4(1), 229. https://doi.org/10.1007/s44217-025-00650-z
29. Saredakis, D., Szpak, A., Birckhead, B., Keage, H. A. D., Rizzo, A., & Loetscher, T. (2020). Factors associated with virtual reality sickness in head-mounted displays: A systematic review and meta-analysis. Frontiers in Human Neuroscience, 14, 96. https://doi.org/10.3389/fnhum.2020.00096
30. Skulmowski, A., & Rey, G. D. (2018). Embodied learning: Introducing a taxonomy based on bodily engagement and task integration. Cognitive Research: Principles and Implications, 3, 6. https://doi.org/10.1186/s41235-018-0092-9
31. Slater, M., & Wilbur, S. (1997). A framework for immersive virtual environments (FIVE): Speculations on the role of presence in virtual environments. Presence: Teleoperators and Virtual Environments, 6(6), 603–616. https://doi.org/10.1162/pres.1997.6.6.603
32. Smith, M. J., Ginger, E. J., Wright, K., Wright, M. A., Taylor, J. L., Humm, L. B., Olsen, D. E., Bell, M. D., & Fleming, M. F. (2014). Virtual reality job interview training in adults with autism spectrum disorder. Journal of Autism and Developmental Disorders, 44(10), 2450–2463. https://doi.org/10.1007/s10803-014-2113-y
33. Stokes, T. F., & Baer, D. M. (1977). An implicit technology of generalization. Journal of Applied Behavior Analysis, 10(2), 349–367. https://doi.org/10.1901/jaba.1977.10-349
34. Sweller, J. (2011). Cognitive load theory. In J. P. Mestre & B. H. Ross (Eds.), Psychology of learning and motivation (Vol. 55, pp. 37–76). Academic Press. https://doi.org/10.1016/B978-0-12-387691-1.00002-8
35. Tychsen, L., & Thio, L. L. (2020). Concern of photosensitive seizures evoked by 3D video displays or virtual reality headsets in children: Current perspective. Eye and Brain, 12, 45–48. https://doi.org/10.2147/EB.S233195
36. Villena-Taranilla, R., Tirado-Olivares, S., Cózar-Gutiérrez, R., & González-Calero, J. A. (2022). Effects of virtual reality on learning outcomes in K-6 education: A meta-analysis. Educational Research Review, 35, 100434. https://doi.org/10.1016/j.edurev.2022.100434
37. Wu, B., Yu, X., & Gu, X. (2020). Effectiveness of immersive virtual reality using head-mounted displays on learning performance: A meta-analysis. British Journal of Educational Technology, 51(6), 1991–2005. https://doi.org/10.1111/bjet.13023
38. Yammine, K., & Violato, C. (2015). A meta-analysis of the educational effectiveness of three-dimensional visualization technologies in teaching anatomy. Anatomical Sciences Education, 8(6), 525–538. https://doi.org/10.1002/ase.1510

End of the guide

Back to geraldajam.com