Gesture Sets
Client:
Study Design:
My Role:
Date
Location
Meta Reality Labs
Mixed Methods Usability Testing
UX Researcher
March 2023
Redmond, WA
Team:
4 member team consisting of 2 UX Researchers, 1 Designer, 1 Research Scientist
Executive Summary
A mixed methods usability study designed to evaluate two competing hand gesture sets (G1 and G2) against physical controllers as a benchmark. The research determined which gesture set best performed across five usability dimensions — ultimately informing a decision to adopt one set and deprecate the other.
BACKGROUND
User Problem & Key Challenges
User Problem: Physical space constraints limited the user's virtual mobility. While physical controllers provided a solution, they diminished immersion and discouraged adoption of available hand tracking.
Team Challenge: 2 sets of hand gestures had been developed and tested. This research effort would ultimately decide on one and potentially deprecate the other.
Research Questions
1. Which gesture set (G1, G2) shows comparable accuracy to controllers (the control group)
2. Which gesture set is favored more by users?
a. through 5 axis: Satisfaction, Intuitiveness-ease of learning, control, ease of use, and sense of efficiency
METHODS
Participants
Target Sample:
Users had to be new to VR with no prior experience. They ranged from 18-65yo
N=31
Although usability studies typically require 5-8 users, our methods included quantitative measures that required at least 25 for statistical soundness
Study Design
A/B Testing
Users alternated between the 2 gesture variables as a starting point to remove familiarity bias followed by controllers (as a benchmark for later analysis).
Quantitative Survey
After users explored the virtual space, they were asked to answer a questionnaire (this was repeated for each variable)
Qualitative feedback
Users were asked why they gave each score and what they enjoyed and disliked about each variable.
Analytics & Presentation
Analytics
I performed repeated measures ANOVA with Bonferroni post hoc correction, and Friedman's Test for ranking. Bonferroni correction ensured statistical validity and strengthened team confidence in results.
Presentation
A slide deck was shown to the team and leadership together
RESULTS
Findings
1. Gesture set 1 (G1) was strongly preferred over the other gesture set (G2) and was comparable to controllers
2. While both gesture sets demonstrated comparable accuracy, G1 intuitive design facilitated user adoption at a faster pace.
Impact
The team applied the more favored gesture set to Meta Quest's hand tracking system by August of that year as a beta and fully released it the following December.
The team explored other applications for G2 with little success
REFLECTIONS
Lessons
Research was introduced late in the product lifecycle, meaning results would ultimately lead to the loss of invested resources regardless of which gesture set was chosen. This taught me to build stakeholder relationships quickly and advocate for research opportunities as soon as they arise. I also dedicated time to learning about exploratory research methods to help teams validate new product ideas before committing resources — a skill I later applied at Eventbrite.
