AnatomyNow

Overview

My team and I were invited to work for The Center of Visualization and Simulation (CVS), located in Da Nang, Vietnam. AnatomyNow is an application for medical anatomy training founded by a startup,

I spent 3 months interning as a UX Designer & Researcher conducting a heuristic evaluation of their iOS app, redesigning the app, and leading usability testing and analysis.

Current state

In the app, users can explore various body systems, create groupings of body parts, interact with 3D models, and toggle between Vietnamese and English. The team of developers that we worked with had never encountered human computer interaction design principles. We began our redesign based on previously accepted design principles while discovering what standards were currently used in other mobile anatomy tools. 

Heuristic evaluation

Navigation

In the original version of Anatomy now, we identified two big issues with the navigation and categorization of objects:

1. The user could only view three items at a time. If the user wanted to view more, they had to keep scrolling until they found the body system they were looking for.

2. The human model on the right-hand side of the navigation screen was not interactive. This is a waste of screen real estate, which is especially important on a mobile device.

Interacting with body elements

In the original version of Anatomy now, we identified three big issues when interacting with body elements:

1. Too many functions

   • Originally, AnatomyNow had 20 different functions. For our redesign, we chose to focus on 9 functions and removed the rest.

2. Icons aren’t clear

   • The icons don’t have a clear signifier (as Charles Pierce said) — meaning there’s not an immediate understanding of what is going to happen or what a user should expect when they interact with one of the 15 icons.

3. Complex user flow

   • If a user goes to remove the frontal bone, they need to select the 3rd icon down to remove it. Okay… didn’t know that icon meant remove, but let’s move on. But what happens when they want to “undo” their action (i.e. bring the frontal bone back)? Instead of de-selecting the original icon that they pressed or having an “undo” feature, the user has to press the 5th icon down.

Grouping body parts

A really cool feature in the application was the ability to create body part groups. However, there was a major flaw: once you created the group, the only way to access it was to go back to the homepage, click on body system, and then in the top right corner click a link for user groups.

So, that’s…

1. Too many clicks

2. Confusing

   • The user has to navigate outside of the area where they just created a group and somehow find their way to the group they just made by trial and error.

Solution

Interacting with body elements

• We decreased the overall icons from 20 to 4 and the interactions from 20 to 11.

• We also labeled icons and added a double tap interaction instead of using an icon.

Sketch

Final UI

Navigation & grouping body parts

• We moved the grouping body parts functionality to the new homepage.

• We increased the number of body systems shown from 3 to at least 8 at a time.

• Decreased the number of steps to view your customized body group by 3.

Final UI

Methodology

Hypotheses

1. The user will navigate faster through the UI using the new search function.

2. User satisfaction scores will be higher in the newer version. 

3. Usability scores will be higher in the newer version.

Materials

To prepare for the experiment the interviewer obtained an iPhone 6 to run the prototypes and a stopwatch application to record the time on task. Both prototypes were created using Axure and had the functionalities of the main page and the skull page. A laptop was used to open the Qualtrics survey that displayed tasks and recorded user input. The interviewer kept a notebook to record time on task and additional observation results. 

Interviewers used a microphone application to record the conversation between the interviewer and the interviewee. The interviewer used a screen recording application on the iPhone to document the interactions that the user had with the prototypes. 

Prototyping software: Axure

Hardware: iPhone 6 and Laptop

Testing Software: Qualtrics (ASQ & PSSUQ), iOS Screen Recorder, and Voice Memos recording app 

Additional items: Notepad

Design (A/B Testing)

Two versions of AnatomyNow created using Axure prototyping software were presented to the participants. The experiment was a between subjects design. The independent variables were the original version and the new version of AnatomyNow. The dependent variables were time on task and user satisfaction.

Participants were randomly assigned to an experimental group, either being exposed to the new or original design. The interviewer assigned a participant ID and submitted it to Qualtrics using the version of the application, the interviewers first initial, and the number of participants that the interviewer has interviewed (e.g. NA1 = new version; Annie; first participant).

Prep for testing

Consent form: Given to participants before user testing to inform them of their rights and the risk of the study. We had to collect their signature and agreement before starting testing.

Interview script: A document for the interviewer to reference during testing and to keep the interviews consistent.

Qualtrics survey: Online survey with prompts to 12 tasks and ASQ based questions after each task. At the end of all tasks questions based on PSSUQ were asked.

Observation log: A document for interviewers to keep track of how many tasks a participant completed, how long it took them to complete the task, and any additional notes.

We also made sure to test the prototype several times before testing to ensure there weren’t any errors during testing.

Uh oh… roadblock

Usability testing planning slammed to a halt when we were told (a few days before going to recruit) that university students were leaving to go on a month long vacation.

Quickly, we met to talk through who we should recruit, when, and where. User testing was starting in a few days and we needed to have a clear plan. We decided to split up and headed to local cafes, the beach, Roll Hostel, and Da Nang Night Market to find willing participants. We channeled our inner sales rep, went out into the field, and cold asked anyone who would listen if they wanted to give feedback on our app.

We also created Tinder profiles to reach a wider net of people. Most of our connections were expats (people who live in Vietnam but are from other countries originally), tourists, and locals who wanted to practice their English.

Usability testing

We had 16 participants — 9 in the control group and 7 in the experimental group. All participants were fluent in English and recruited from cafes, the night market, and online.

How it worked

Before testing, we obtained written consent from the participant and answered any questions they had before starting the test. Once we were good to go, we started up the Axure RP prototype and let them know that they would be asked to complete a couple of tasks using the iPhone. After each task they’d be asked to complete a couple questions using the laptop (through Qualtrics).

A participant could forfeit a task if it was too challenging at any time. Once the participant let us know that they’d completed the tasks or they felt they couldn’t complete it, they’d answer questions based on ASQ. After they completed all tasks they answered a series of questions based on PSSUQ.

At the end of the experiment, we gave them a verbal debriefing and thanked them for their participation.

 Results

We analyzed:

1. Usability: Measured by self-reported survey (PSSUQ).

2. Efficiency: Measured by the amount of time a user took to complete a task.

3. Effectiveness: Measured by the rate that users completed the task.

We used SPSS and JASP to analyze our results.

• Task Completion Rate: 98%, a 36% increase from the original version

• Time on Task: 6.6s, a 74% decrease from the original version

• PSSUQ: 6.7 out of 7, a 109% increase from the original version

• PSSUQ p<0.05, achieved significance in academic standards

Final thoughts

I learned a lot from my internship:

• How to work with an international multi-disciplinary team

• How to advocate for my ideas to a larger group & collaborate to create more diverse ideas.

• Things can change quickly so be able to adapt quickly.

• Your designs won’t always be validated, and that’s okay — be ready and willing to make adjustments or completely rethink the problem and test again.

• Networking is so important!

• Vietnam’s coffee is unmatched and I’m going back asap.

Other cool opportunities:

• Gave a talk about how to empower and support young girls to get into and stay in technology at the SURF Conference (a start-up conference) with Tara O’Grady.

• Gave a talk on the pervasive nature of augmented reality in education at a Google I/O Extended event with Tara O’Grady.

 Meet the team

Shoutout to Kristen, Oliver, and Tim for making our experience fun (we love a good karaoke night), giving valuable honest feedback, and being able to adjust on the fly! This was such an incredible opportunity and I’m so grateful that I had a wonderfully supportive team.