XR Design
UX Research
Usability Testing
VR Learning Module for Midwives for third stage of Labor
Making a VR learning module usable for nurses and midwives who had never held a headset
The module trained midwives in managing childbirth complications. It worked for anyone fluent in tech. Our actual users, nurses in government institutes with little English and no VR exposure, could not locate themselves in the environment.
Identified the usability gap and argued it onto the roadmap. Redesigned the failing interactions. Co-designed and shipped a new onboarding module with an XR designer-developer.
Contents
1
The project was funded by UNFPA to launch VR labs in National Midwifery Training Institutes in North India.
2
Tracking the success of the module for scalability to other cities or even other countries.
3
Discover
01
The first question that I had was
“If this module is made for nurses and midwives in smaller towns of India, shouldn't we test on people who are not aware of navigating through VR environments?"
Secondary Research
After our first research round, I found the following about VR
of the users are not familiar with VR devices or environments
of users start to feel dizzy while using VR
non-technical users require immediate, hands-on guidance to figure out the interfaces
Initial Testing
We partnered with Fernandez Hospitals, for medical guidance and product development. We ran out initial testing with a few midwives at Fernandez and gathered data regarding usability and user behaviour to enhance the module.
We studied the behaviour and module usability through 12 participants over 2 days. We have asked each participant to re-visit the module to understand information retention rates.
Additionally, designed these visual learning aids for users to learn the names of the buttons and hand placements.
Define
02
midwives/ nurses of varying levels of medical expertise in phase 1
participants completed the module with guidance on the first try
participants needed guidance throughout the module on the first try
participants did not require guidance during the second round
A.
Orientation & onboarding
Users did not understand the concept of VR and that severely affected the way they moved around. Many users did not locate the objects in heir vicinity, they had trouble moving around in the environment and most importantly they did not understand how to perform basic actions such as picking up the object.
Few users needed constant reminders that movements in physical spaces are mimicked in the VR environment.
Directing the users to look at their surrondings
Guiding the users about which button should be used
B.
Interaction design
Few interface related issues were observed during the testing phase, such as:
Users double clicking on the next button which led to skipping steps
Users not being able to see which button need to be used
Accessibility issues and button states discripencies
C.
Feedback & Communication
Users did not have enough audio cues, success cues or messages or textual cues to fall back after an instruction.
D.
Cognitive & behavioural design
Many participants taken a route through the module in the way the design was not intended.
E.
Clinical
accuracy
Discrepancies between the simulation and real-world labour room practice, flagged directly by practicing midwives.
Develop
03
We designed a four-room gamified onboarding experience. Each room introduced one layer of VR interaction, building on the last.
Additionally, designed these visual learning aids for users to learn the names of the buttons and hand placements.
midwives/ nurses of varying levels of medical expertise in phase 2
participants completed the module, unguided on the first try
participants completed the module, unguided on the second try
was observed to be persistent failure point across both test rounds
Deliver
04
Disclaimer
Sensitive visuals ahead
Interface redesigns
In step 4, Controlled cord traction, users are supposed to learn how to analyze a contraction using their left hand pull the umbilical cord with their right simultaneously when the contraction is occurring.
Step used proximity grab first, then required a separate grab button with no indication of the difference.
Users gave up assuming it was a bug. Here were a few fixes I suggested.
BEFORE
Green to red
AFTER
