XR / Spatial Computing

A 10-Minute VR Walkthrough Beat 27 Slides at Explaining Height Barriers — With Effect Sizes Around g = 2

100 participants saw either a first-person Unity walkthrough or a silent self-paced slide deck covering the same events. Narrative engagement and state empathy were both far higher for the video. The authors are careful to say what that does not prove.

Big World, Little Me is a 10 minute 52 second first-person walkthrough recorded from a Unity-based VR environment. It depicts barriers encountered by people with dwarfism and other forms of short stature: domestic reachability, public transport, and employment.

The evaluation was posted to arXiv on 22 October 2026 (arXiv:2610.30731) by Chan In Sio, Alex Mann, Qian Ming Jia, Xuecheng Guo, Yan Zhuo, Xinyu Ning, Kaikuo Liao and Lik-Hang Lee. The Comments field reads: “14 pages, 2 figures. Accepted for presentation and publication at ICVRT 2026. Author manuscript.”

We checked the venue rather than assuming it had happened. ICVRT 2026 — the 9th International Conference on Virtual Reality Technology — runs 25–27 December 2026 in Harbin, China, with proceedings in the ACM International Conference Proceedings Series. So this is accepted work whose venue is still two months away, which means there is no DOI to cite yet; the arXiv record is the thing to reference, and the PDF is an author manuscript rather than the published version.

The comparison

A non-randomised between-subjects study, 100 participants aged 15–48. Each saw one of two things covering the same principal events:

  • the fixed-duration audiovisual walkthrough (10:52), or
  • a silent, self-paced 27-slide presentation.

Results:

MeasureWalkthroughSlidesDifferenceEffect size
Narrative engagementM = 5.54, SD = 0.51M = 4.24, SD = 0.731.30, 95% CI [1.05, 1.55]Hedges’ g = 2.05
State empathyM = 3.15, SD = 0.33M = 2.21, SD = 0.600.94, 95% CI [0.74, 1.13]Hedges’ g = 1.91

Both at p < .001. Exploratory supportive-intention ratings rose more in the walkthrough condition. Participants also rated the walkthrough as more dynamic, continuous, spatially clear and information-rich.

Those effect sizes are very large. For context, a g around 0.8 is conventionally “large”; 2.05 means the distributions barely overlap. Confidence intervals that tight on n = 100 are consistent with a real and substantial difference rather than a marginal one.

The authors’ own caveats, which are the most useful part

This paper is unusually disciplined about what it has shown, and anyone citing it should carry the limitations along with the numbers.

It does not isolate a mechanism. In the authors’ words, the findings “do not isolate the effects of motion, continuity, pacing, information quantity, or participant control”. The two conditions differ on all of those at once. A continuous narrated video versus a silent self-paced deck differs in so many ways that “VR works better” is not a conclusion the design supports — it could be the motion, the audio, the enforced pacing, or simply that a video is harder to skim than slides.

It is not immersive VR. Stated outright:

The walkthrough should be understood as screen-based VR-authored media rather than immersive VR or a reproduction of lived experience.

A video recorded from a Unity scene, watched on a screen. No headset. The “VR” in the title describes how the content was authored, not how it was consumed — which makes the result more portable, since a video needs no hardware, but means it says nothing about head-mounted displays.

It is not lived experience. That phrase is doing deliberate work. There is a long-standing critique of “empathy machine” claims for VR disability simulations: that briefly experiencing a simulated barrier can produce confidence about understanding a life it does not resemble, and can centre the non-disabled viewer’s feelings. The authors decline the stronger claim.

It was not randomised. Non-randomised assignment leaves open that the groups differed systematically, and with a self-selected mechanism that is not a small concern.

What it is good evidence for

Taken narrowly, it is solid evidence for something practical: if you need to communicate a spatial accessibility barrier to people who have not encountered it, a short authored first-person walkthrough communicates more engagingly than a slide deck, and you do not need a headset to deliver it.

That is directly applicable. Accessibility arguments are usually made in planning documents, consultation packs and training slides — the slide condition is not a straw man, it is the status quo. A Unity scene rendered to a ten-minute video is within reach of a small team and distributes as a file.

The honest framing for using it: this changes how an argument lands, not what the argument is. The barriers the piece depicts — fixed-height environments, and what the paper calls “discriminatory social assumptions” — are documented independently of whether a video about them is engaging.

Sources