User Experience Research · Spatial Computing · AR/VR/MR

The environment is part of the interface

The work here asked a researchable question: when information lives in three dimensions around a person, what changes in how they understand it, feel it, and act on it. Every answer on this page came from a controlled study.

The same question ran through every project

Across a virtual operating room, a study pitting spatial overlay against print, a retail presence experiment, eye-tracking with practicing radiologists, and a survey of working designers, one theme kept surfacing: the space around an interface changes what people understand, feel, and do. 

The work runs in sequence. It starts by building an immersive environment people could learn inside, tests whether spatial media outperforms conventional media on the same content, isolates the mechanism that makes it work, follows attention down to the movement of the eye, and closes by asking why design teams are slow to adopt tools that demonstrably help.

Build​

VR Operating Room: Immersive Visualization Lab

Comparison

Spatial overlay vs flat media: Print / AR / mixed reality

Mechanism

Presence mediates UX: HMD and desktop VR experiments

Attention

Eye-tracking in clinic: Healthcare technology

Adoption

Why designers don't adopt: Field study, 54 practitioners​

Virtual Operating Room

New residents and operating-room staff had to learn an unfamiliar, equipment-dense space from detached materials: manuals, lectures, static diagrams. None of those convey the thing that actually matters in an OR, which is where everything sits in relation to everything else, and to user.

Co-created with the University of Missouri School of Medicine, this simulator turned room layout, equipment, and procedural workflow into something staff could practice inside. Stereoscopic 3D for depth. Controller-driven navigation for movement. The space itself did the teaching.

Stereoscopic dual view of the virtual operating room, captured in the headset

Fig. 01 · Stereoscopic dual view of the operating room, captured in the headset.

For creating accurate virtual operating room, all details measured and recreated digitally

Fig. 02 · Room and equipment measured and rebuilt digitally to scale.

A functional prototype that ran on real hardware.

A working, navigable environment used in sessions with real clinical staff.

Operating-room workflow modeled in 3D.

Bed, lights, anesthesia cart, monitors, and supply zones positioned to procedural reality and modeled to scale.

Embodied interaction.

Controller-driven navigation let users travel the room and build its spatial relationships firsthand.

Role: 3D modeling, in-VR UI, and interaction design.

Co-created the simulator inside the Immersive Visualization Lab.

Spatial overlay outperformed conventional media for product information

When someone evaluates a product, the information that shapes their judgment arrives through whatever medium happens to carry it: a printed sheet, a tablet screen, or content anchored in space around the object itself. This study tested all three with the same product and the same content, in a three-condition within-subject design: print, tablet AR, and head-mounted mixed reality.

The media split along different strengths. Print drove the highest cognitive involvement. The spatial condition, where information was overlaid and anchored to the product in the user’s real field of view, was rated the most informative of the three. The medium was not neutral: where the information lived changed how well it landed.

Sustainability information overlaid and anchored to a product in head-mounted mixed reality

Fig. 03 · Sustainability information anchored to a product in head-mounted mixed reality.

PRINT MEDIA

Highest cognitive involvement

SCREEN AR

Information overlaid on screen

HMD MR

Rated by users as most informative

Three-condition within-subject design.

Every participant experienced all three media, so the differences trace to the medium itself.

Print vs. tablet AR vs. head-mounted MR.

Same product, same sustainability content, three delivery formats.

Anchored information landed best; print worked hardest.

Head-mounted mixed reality was rated the most informative format, while print produced the highest cognitive involvement.

Role: ran the study and co-built the 3D content.

Conceptual bridge

This connects to a deeper question. If the medium changes response, why? The next study isolates the mechanism.

Presence is the mechanism that makes spatial media work

Spatial media changes user response, and the design question is the mechanism: what is the lever. This study ran two parallel experiments, one in a fully immersive head-mounted display and one in desktop VR, using the same high-fidelity retail environment in both. The model under test: attention shapes user experience, but indirectly, through a user’s sense of presence, the felt sense of being there.

Across both display types, presence partially mediated the effect of attention on experience. Immersion that deepened the felt sense of being there improved the experience, and attention kept a direct effect of its own. This is the empirical core of the doctoral research: the environment is a working part of the interface.

Participant at a workstation wearing a head-mounted display during a VR retail experiment
Participant at a workstation playing with joystick and 2.5D VR retail experiment

Fig. 04 · Parallel head-mounted and desktop VR retail experiments.

Two experiments, two display types.

Fully immersive HMD (stereoscopic 3D) and 2.5D desktop VR, to test whether the finding held across platforms.

Presence mediates attention's effect on experience.

Confirmed by regression analysis and Sobel test in both studies.

Role: designed, built, and ran both studies.

Mediation model

Presence as the mechanism, measured across two studies
Relationship among sense of presence, attention, and user experience in the virtual environments
Fig. 05 · Mediation model: attention, presence, user experience. a · X→M b · M→Y c · X→Y total c′ · X→Y direct (with M)

STUDY 1

N = 90
Immersive HMD · stereoscopic 3D

c · X→Y total

b = .418 · t = 5.29 · p < .001

c′ · X→Y direct

b = .30 · t = 4.75 · p < .001

b · M→Y

b = .248 · t = 2.28 · p = .025

Sobel

z = 2.19 · p = .029 · partial mediation

Reliability α

UX .89 / Pres .96 / Att .75

VIF

1.06 · no multicollinearity concern

STUDY 2

N = 91
Desktop VR · 2.5D, large display with game controller

c · X→Y total

b = .32 · t = 3.74 · p < .001

c′ · X→Y direct

b = .20 · t = 2.83 · p = .006

b · M→Y

b = .51 · t = 7.15 · p < .001

Sobel

z = 2.21 · p = .027 · partial mediation

Reliability α

UX .92 / Pres .96 / Att .84

VIF

1.06 · no multicollinearity concern

Recap

Attention improved experience in both studies, and part of that effect ran through presence. The direct path stayed significant too, so the mechanism is shared: presence carries a real portion of the effect, and attention still does work of its own.

Eye-tracking turned gaze into design decisions

On a large healthcare technology partnership building a next-generation medical imaging platform, interviews could only carry the question so far. The real question was where clinicians actually looked: what they noticed, what they missed, what slowed them down, and where the interface created confusion.

The study put eye-tracking on practicing radiologists working through realistic reading tasks, and ran the analysis on measured gaze: dwell time, fixation count, and time to first fixation across areas of interest on each interface. Attention was measured at the level of the eye, then translated into changes a product team could act on.

Eye-tracking gaze visualization over a clinical reading interface, details withheld for confidentiality

Fig. 06 · Gaze visualization over a clinical reading interface, details withheld for confidentiality.

Attention measured at the eye.

Eye-tracking captured real gaze behavior from practicing radiologists, with self-report kept as a cross-check rather than the evidence.

Within-subject study with layered methods.

Quantitative usability testing with measured gaze metrics, qualitative gaze-replay analysis, and pre and post task satisfaction surveys.

Findings became concrete design changes.

Each result mapped to a specific change: reordering information, shifting emphasis, fixing a confusing control, adjusting contrast for readability.

Role: designer and researcher.

Planned the study, ran the sessions, analyzed the data, and turned the findings into recommendations the product team could execute.

Eye-tracking results

Attention measured directly, with practicing radiologists

ANALYSIS OF VARIANCE

Measured gaze across AoI, Tukey HSD post-hoc

Interface A · AOI importance

F = 10.85 · p < .001

Interface B · AOI importance

F = 8.11 · p < .001

Attention was not uniform on either interface: some regions drew far more gaze than others.

CROSS-METHOD CHECK

Measured gaze versus what users reported

Gaze vs self-reported noticeability

chi-square · p = .025

Gaze behavior was tested against self-reports, and the gap was significant on at least one element. Eye-tracking caught what surveys alone would miss.

Study scope and methods

PARTICIPANTS

10 radiologists in quantitative testing, 7 in qualitative gaze-replay
 

SCOPE

5 radiology specialties, two experience groups: 5 to 9 years and 15 or more

EYE-TRACKER

30 to 60 Hz sampling, 0.5 to 1 degree accuracy

ANALYSIS

ANOVA, Tukey HSD, t-test, chi-square, scanpath protocol, heatmaps. Satisfaction by SUS and CSUQ

Gaze replay revealed how each interface was actually read

Mirrored F-shape on one surface

Gaze concentrated along the top and the leading edge, then dropped off: a predictable scan path that tells you exactly where to place what matters most.

Segregated reading on another

Attention split into separated zones instead of a single path, a sign the layout was fragmenting the read and forcing extra work.

Recap

Radiologists analyzed and synthesized one interface faster than the other, and the gaze data explained why. Eye-tracking mapped where attention concentrated, where it broke, and where the layout could be reconnected for faster, more accurate reads. Each finding became a specific, executable change for the product team.

A field study measured why design teams underuse tools that help

Building immersive prototypes and proving they change user response raises an obvious follow-up: if the tools help this much, why are design firms slow to adopt them.

This mixed-methods study, run at the 80-plus-person office of a 1,200-employee international design and engineering firm and published at the International Conference on Engineering Design, went looking for the barriers. A focus group surfaced them; a survey of 54 practitioners measured them.

Mixed methods.

A focus group of six practitioners surfaced the barriers; a survey of 54 measured them.

The barriers operate as one system.

Designer perception, manager attitude, and fear of learning the technology under workload pressure proved interrelated with funding, training, and technical support. The study's recommendation: start by resolving the internal barriers.

Experience and willingness did not predict attitude.

A regression found neither years of experience nor willingness to learn was a significant predictor of adoption attitude within the firm.

Role: Faculty advisor on the graduate research project and co-author.

Published at International Conference on Engineering Design, Delft

Part A

Percentage of practitioners comfortable producing each output

360° still renders
31.7%
Walk-through animation
28%
VR production
21.5%
AR production
11.2%
MR production
7.6%

Practitioner comfort by output type. The more spatial the medium, the fewer designers feel comfortable producing it.

Part B

Where the barriers live

Internal Factors

Designer perception of VR, AR, and MR​

Manager attitude toward integration in design service​

Perception and managerial framing, not hardware, set the ceiling on adoption.

 

External Factors

Hardware and software availability and reliability

Training programs

Funding

Managerial support

In-house champions

Technical support

Regression finding

 

Years of experience was NOT a significant
predictor of adoption attitude.

F(2, 25) = 0.62 · p = .55

 

Within the firm, seniority told you nothing about who would adopt.

What this establishes

Taken together, these projects form one argument: spatial context is a functional part of the interface. The work proved it constructively, by building a clinical simulator people learned inside, and empirically, by showing spatial overlay outperforming flat media on the same content and by isolating presence as the mechanism that makes immersive media work.

It carried the same question down to the movement of the eye, turning what radiologists looked at into concrete interface changes. The adoption study then turned the lens on the field itself, naming why teams underuse tools that demonstrably help. This is the foundation the rest of the portfolio builds on: a designer who treats the environment, and attention itself, as variables to be tested, and who pairs prototypes with the evidence that justifies them.

What it protects and enables

This research capability catches attention and comprehension failures before they ship, and turns interface debate into measured decisions. The methods that validated these spatial interfaces carry directly into regulated devices and into software used at large scale. That is the throughline of the work: product design and human factors, from medical devices to products used by millions.

What it costs to do properly

Controlled studies take time, real participants, and honest analysis. It means running accessible conditions next to high-fidelity ones, and being willing to let the data overrule the prettier design. Done properly, it replaces opinion with evidence at the point where a decision is expensive to reverse.

TOOLS

SolidWorks · 3ds Max · Unity · Oculus · HoloLens · HTC Vive · Mobile and desktop eye-trackers

IMPACT

Peer-reviewed work in the Journal of Product & Brand Management · ACM Web3D Conference · International Conference on Engineering Design (ICED), and more

COLLABORATORS

Immersive Visualization Lab · U of Missouri School of Medicine · U of Missouri Information Experience Lab · U of Minnesota VR Lab