# Why these three studies

Thanos asked for XR experiences he can use to run research and publish, without knowing which topics are current.
This note records what was looked up (November 2026), which topics were chosen, and why each one suits a small studio
with no headset budget. **Verify every reference before citing it in a paper**: they were checked for existence and
gist, not read in full.

## What the field is working on

- **IEEE VR 2026 programme** ([ieeevr.org](https://ieeevr.org/dev/program/papers)): the largest clusters of sessions are
  perception (about 7 sessions), avatars, embodiment and presence (about 6), haptics, locomotion and redirection, and a
  dedicated cybersickness session. Training and education, accessibility and social VR form smaller clusters. Example
  papers: *Peripheral Teleportation: A Rest Frame Design to Mitigate Cybersickness During Virtual Locomotion*;
  *Reduction of Motion Complexity as an Objective Indicator of Cybersickness*; *Misperception of the distance of virtual
  augmentations*; *Implicit Learning of Professional Skills through Immersive Virtual Reality: a Media Comparison Study*.
- **How experiences are evaluated** (Li, He & Kristensson, *How Do We Evaluate Experiences in Immersive Environments?*,
  CHI 2026, [arXiv 2601.17811](https://arxiv.org/abs/2601.17811)): a review of 375 papers from CHI, UIST, VRST, SUI,
  IEEE VR, ISMAR and TVCG. 85.6 % use questionnaires and 62.7 % task performance, but instruments vary widely; the
  authors call for **standard instruments, open and shareable protocols and datasets**. A kit that ships its full
  protocol, standard questionnaires and an export format answers that call directly.
- **Remote and unsupervised XR studies** are an accepted method with known pitfalls (Mottelson, Petersen, Lilija & Makransky,
  *Conducting Unsupervised Virtual Reality User Studies Online*, Frontiers in Virtual Reality 2021,
  [doi:10.3389/frvir.2021.681482](https://doi.org/10.3389/frvir.2021.681482); *CrowdXR: Pitfalls and Potentials of
  Experiments with Remote Participants*, [PolyU record](https://research.polyu.edu.hk/en/publications/crowdxr-pitfalls-and-potentials-of-experiments-with-remote-partic/)):
  less control over the setting, more diverse participants, and the need to log the device and frame rate. Browser-based
  WebXR studies need no app store and no installation, which is what makes them practical for a small lab.

## What a small lab can do well

No headset budget, no motion-capture room and no backend means: studies that run in a browser, on participants' own
headsets, phones or laptops, or on a few borrowed headsets in a lab; measures that need no hardware beyond the device
(questionnaires, responses, task logs, head pose, frame rate); and comparisons where the *device* is recorded as a factor
rather than controlled. Each study below works on a VR headset and on an ordinary screen, so it can be piloted with
laptops and phones today and run with headsets when a partner or a funded project provides them.

## Study 1: Comfort Ride (cybersickness mitigation)

**Question.** During passive virtual travel, does a dynamic field-of-view restriction or a rest frame reduce
cybersickness compared with no aid, and does the answer differ between headsets and screens?

**Why it is publishable.** Cybersickness has its own session at IEEE VR 2026 and remains the main barrier to VR
adoption. FOV restriction (Fernandes & Feiner, *Combating VR sickness through subtle dynamic field-of-view
modification*, IEEE 3DUI 2016) and rest frames (Prothero's rest-frame hypothesis, 1998; Cao, Jerald & Kopper,
*Visually-induced motion sickness reduction via static and dynamic rest frames*, IEEE VR 2018) are the two most used
aids, but are rarely compared head to head under one stimulus, and almost never on ordinary screens, where most people
now meet 3D travel. A fixed, logged stimulus (route, speed profile and frame rate saved with every response) makes the
study replicable, which is what the CHI 2026 review asks for. Stanney, Lawson, Rokers et al. (*Identifying causes of and
solutions for cybersickness in immersive technology*, IJHCI 2020) set the agenda this fits.

**Design.** Between subjects (sickness carries over, so each person sees one condition). Measures: SSQ (Kennedy et
al., 1993) before and after, Fast Motion Sickness scale every minute (Keshavarz & Hecht, 2011), time to stopping,
IPQ presence (Schubert et al., 2001), frame rate. Covariates: travel-sickness history, gaming, VR experience, device.
**Sample size:** for three groups and a medium effect (f = 0.25, α = .05, power .80) about 159 people in total (53 per
group); a pilot with 10 per group first.

**Possible venues:** IEEE VR or ISMAR posters for a pilot; IEEE VR / VRST papers or Frontiers in Virtual Reality for the
full study.

## Study 2: How Far (distance perception in VR, AR and on screens)

**Question.** How much do people underestimate the distance of a virtual object in VR, in phone or headset AR and on a
screen, and does a contact shadow reduce the error? Does a textured ground help?

**Why it is publishable.** Perception is the largest cluster at IEEE VR 2026, including a paper on misperceived
distances of AR content. Distance underestimation in VR is well documented (Renner, Velichkovsky & Helmert, *The
perception of egocentric distances in virtual environments: a review*, ACM Computing Surveys 2013), shrinks with newer
headsets (Kelly, *Distance perception in virtual reality: a meta-analysis of the effect of head-mounted display
characteristics*, IEEE TVCG 2023), and shadows and ground contact change it in AR (Diaz, Walker, Szafir & Szafir,
*Designing for depth perceptions in augmented reality*, ISMAR 2017; Adams et al., *Shedding light on cast shadows*,
IEEE TVCG 2022). Few studies compare **the same task on consumer VR, phone AR and screens** with participants' own
devices. That cross-device data set is the contribution.

**Design.** Within subjects: shadow (on/off) × 5 distances × 2 repetitions, random order, after 2 practice trials.
Between subjects: textured vs plain ground (VR and screens). Device is recorded (a quasi-experimental factor). Measure:
verbal estimate in metres (typed); ratio to the true head-to-object distance; head movement. **Limitation to state:**
verbal reports underestimate more than blind walking, which cannot be run remotely. **Sample size:** 24 to 30 people per
device type detect a medium within-subject shadow effect.

**Possible venues:** IEEE VR / ISMAR (perception tracks), ACM SAP (Symposium on Applied Perception), IEEE TVCG.

## Study 3: Pump Start-up (training transfer with in-world guidance)

**Question.** Does highlighting the next control in the world during VR procedure training help while training but hurt
recall when the help is gone (the guidance hypothesis)?

**Why it is publishable.** Training and education are a steady IEEE VR theme (2026: a media-comparison study of
professional skills) and the core of Vourla Labs' business. In-situ instructions cut errors during assembly (Tang et
al., *Comparative effectiveness of augmented reality in object assembly*, CHI 2003), but motor-learning research
predicts that constant guidance can hurt later unaided performance (Salmoni, Schmidt & Walter, *Knowledge of results and
motor learning*, Psychological Bulletin 1984). Immersion can raise presence without raising learning (Makransky,
Terkildsen & Mayer, *Learning and Instruction* 2019). Industrial VR training is widely studied (Gavish et al.,
*Interactive Learning Environments* 2015), yet cue-versus-text comparisons that **measure unaided recall and delayed
retention** are rare. The answer matters directly to anyone building VR training.

**Design.** Between subjects: cue (step list plus an in-world marker on the next control) vs text (step list only).
Two guided runs, then a test from memory with no help or feedback; an optional delayed test with the same code
(`?phase=retention`). Measures: errors, time, steps done in order, correct steps before the first error, NASA-TLX
(raw), SUS, IPQ. **Sample size:** two groups, medium effect (d = 0.5): 64 per group; large (d = 0.8): 26 per group.

**Possible venues:** IEEE VR / ISMAR (training and education), CHI late-breaking work, *Computers & Education*,
*Virtual Reality* (Springer), Frontiers in Virtual Reality.

## Topics considered and not built (yet)

- **Avatars and embodiment** (the second-largest IEEE VR cluster): needs full-body or hand tracking and a headset for
  every participant; a later study once headsets are available. The kit's hand support (lab/26) is a starting point.
- **Haptics:** needs hardware.
- **Accessibility:** a strong topic (low vision, motor impairments), but it must be designed with the participants'
  community, not for them; a good collaboration with a university group.
- **Social VR and co-presence:** possible with the live-session tools in `shared/live-session.js`, but harder to run
  remotely in a controlled way.

## Publishing well

Preregister each study (OSF, aspredicted.org) with the hypotheses, sample size, exclusions and analysis before data
collection; publish the data (anonymous) and this code with the paper; report device, browser and frame rate; run a
pilot first and fix the protocol before the real study. A university partner brings ethics approval, participants and
co-authors: see ETHICS.md.
