VR (extended reality) for skills training

Bateman (2005)

Why train skills in XR

• train the untrainable

• more effective than conventional training

training the untrainable

• practically (astronauts) & theoretically untrainable (warzones)

  • issues regarding safety, difficulty & cost

how could XR be more effective than conventional training

• Combined technologies can provide performance feedback

• wow factor boosts engagement

• Everything is manipulable

measuring performance of skills

• interested in the transfer of skills to physical reality

• almost all CR training is indirect

good XR research should measure?

• learning (how much skills has improved in XR)

• transfer (how the training translates to physical reality)

when should we use XR for training 

• Feasible 

  • Can training be done without XR, is it expernsive, will the target group benefit

• transferable

  • can training transfer to the target skill, is it more successful

ideal process of training studies

1. baseline measures

2. blind participants randomisation 

3. different levels (XR, conventional, control)

4. post training & transfer test

5. retention test

probability of skills transfer in XR research depends on

• presence

• validity

• individual differences

presence in XR skills transfer

• immersion

• ease of interaction

• personalisation 

validity in XR skills transfer

• external validity

  • enviromental matching

• internal validity

  • thematic consistency

• physical validity

  • reality physics

individual differences in XR skills transfer

• XR experience

• cybersickness

• experience

• skills experience 

things which must be considered during XR research

• Exclude P w/high likelihood of experiencing cybersickness (e.g., previous history with similar technology)

• Measure any outcome-relevant variables at baseline

• Measure cybersickness symptoms throughout procedure (using biomarkers & self-report)

• Measure transfer & task performance at baseline and post-training Include an active control condition

• Measure presence & cognitive load throughout the procedure If possible, measure skill retention at follow-up

Is XR training actually any good (Bateman, 2005)

• metanalysis which used RCT that tested XR training of psychomotor skill

• passive controls (no training, waiting list)

• active controls (convectional training)

findings of Is XR training actually any good (Bateman, 2025)

• Consistent negative effect (not stat. sig.) of haptic feedback on psychomotor skill learning

• all models found significant positive effects of XR training over non-XR training on task performance 

  • XR more effective than other training interventions

• No clear relationships between methodological features and psychomotor skill learning

  • AR, VR

  • Type of skill (surgical, sport, music, etc.)

what factors predict learning 

• Conducted Bayesian Network analysis

role of haptic feedback

• negative feedback on learning

• haptic was too broad 

XR is not automatically superior in training

• its value depends on cost–benefit trade-offs for the specific skill and context

• it has to beat what you already do (sim labs, supervised practice, on-the-job training), not just “no training”

• worth it when it unlocks things you can’t do cheaply or safely otherwise

Bateman et al (2025)’s recommendations for XR training

• Better, more standardised RCT designs (dose, timing, controls).

• Routine pre-registration and full methodological + technical reporting.

• Systematic measurement of cybersickness and other confounds.

Makransky & Petersen – CAMIL (2021)

• VR is useful when you deliberately use methods that channel presence/agency into learning & manage cognitive load

• otherwise it can be flashy but inefficient

  • IVR can increases extraneous cognitive load compared w/flat-screen/multimedia if badly designed