Project Overview

The Institute for Health Metrics and Evaluation (IHME) is an independent global health research center at the University of Washington funded primarily by the Bill and Melinda Gates Foundation.

The Introduction to the Global Burden of Disease (GBD) is IHME’s flagship online course serving scientific collaborators, journalists, and policymakers.

I was hired in January 2021, at the height of the pandemic, to reimagine the Introduction to the GBD. Part of my role was to establish a UX design process on the learning and development team.

Problem

IHME’s work modeling the COVID-19 virus led to exposure on CNN, BBC, NPR, The New York Times, Twitter, and many other mainstream media outlets, attracting large audiences to their data tools and studies. As a result, there was an urgent need to educate the public on how to use IHME’s tools, interpret the data, and build trust through transparency in the modeling process.

Audience and Learner Goals

• 7,500 existing scientific collaborators in 157 countries
• Researchers and scientists need to understand the modeling process as an entry point to collaboration.
• Journalists need to understand the GBD study so they can report timely, relevant data.
• Government health officials need unbiased, reliable data to support policy decisions.

Organizational Goals

• Achieve IHME’s mission to improve the health of the world’s population by making available the highest-quality information on population health, its determinants, and the performance of health systems.
• Reduce criticism from peers regarding a lack of transparency in IHME’s data modeling process.
• Grow IHME’s scientific collaborator network.
• Educate journalists and policymakers on how to use and understand IHME’s data tools.

My Responsibilities

Full scope design and development including instructional design, user experience design, UI/graphic design, motion graphics, data visualization, eLearning development, project management, and localization.

Process

The first step was to meet with stakeholders, review survey data, and evaluate existing content. This established organizational goals, expectations, timelines, and an overview of subject matter. We had 19 scientific articles, primitive infographics, and outdated data. Learner feedback indicated existing content was overwhelming and poorly targeted.

We agreed on three months to design and develop the training. The next step was to release a minimum viable product as quickly as possible. We evaluated existing content for criticality, streamlined, synthesized and agreed on a high priority content outline for our first iteration.

In three weeks, I delivered six courses that served as prototypes to test with a small subset of our public audience and stakeholders. (~200 learners)

With a minimum viable product in place, we were ready to interview users. These interviews would inform my investment in development time going forward.

User Research

With the stage set for user research we conducted four, 45 minute user interviews via Zoom to empathize with learners and come to a deeper understanding of their goals, pain points, and needs. I worked around the clock to accommodate users in Iran, Portugal, Greece, and New York.

Sharing Insights

After working together to pull quotes from user interviews and affinity map themes, I shared user stories and research insights with stakeholders. I chose to share the qualitative data points all four users shared in common. Qualitative data was further validated by survey responses. If all four users had the same requests, and those requests were present in survey data, this would be a reliable starting point for improving content. The requests were for engaging multimedia, interactivity, context on how to collaborate, and more realistic data visualization tool training.

Competitive Analysis

I looked at how other organizations were approaching similar problems. IHME is unique and the Introduction to the GBD has more in common with customer education than internal workplace training. I looked at big data/data viz companies like Tableau and ESRI, and organizations who present public health data like Oxford’s Our World in Data and John Hopkin’s School of Medicine.

While Tableau and ESRI provided valuable insights on how to teach data tools, Oxford and John Hopkin’s provided a more journalistic approach. IHME sits comfortably between these organizations, and that balance guided my approach to content design and presentation.

Heuristic Evaluation

I tested concepts and user flows, evaluated information architecture, and evaluated content heuristically, applying Jakob Nielsen’s 10 Principles for Interaction Design, and Mayer’s Cognitive Principles of Multimedia Learning.

Heuristic evaluation revealed that the learning management system (LMS) wasn’t ready for production. It was set up for staff, and it didn’t receive the same treatment as external products. I conducted a quick round of thirty-minute observational usability tests with the LMS admin and five learners.

It was a challenge to balance the needs of staff with the desired user experience for our new external audience. Our solution was to split the LMS into internal and external sites. To external users the LMS now appears as a simple card gallery of Articulate Rise courses. As an added benefit, we were able to improve the user experience for internal staff by removing some distractions.

Second Iteration

With a renewed understanding of learner goals, pain points, and needs, I revised the content outline and identified key areas for multimedia, interactivity, and context. We added an assessment/certificate, and a case study provided by a GBD Research Scientist. I streamlined and synthesized the content from two courses to eliminate extraneous information.

Development

This is an overview of the process I followed to design and develop one media asset in the Intro to GBD Online training. There are 50+ media assets in the program, spanning six courses.

Tools

• Asana for project management
• Google Suite for content collaboration
• Microsoft Office for content collaboration, scheduling, and communication
• Zoom for international user interviews and collaboration
• Miro for design thinking exercises
• Adobe XD for prototyping
• Figma for prototyping and user flows
• Excel for data preparation
• Adobe Illustrator for cleaning and separating chart layers, and for infographics
• VS Code for editing .KML GIS data in prep for Google Earth Studio
• Google Earth Studio for choropleth map overlays
• Adobe After Effects for data animation and motion graphics
• Articulate Rise for content delivery and assessments
• Survey Monkey for learner data collection

Deliverables

These are screenshots and outtakes from the training. The courses are behind a paywall for the general public, and free for scientific collaborators who’ve registered with IHME’s collaborator network.

Six Courses

Data Visualization and Motion Graphics

Impact

Initial Learner Reaction

• “Did this course improve your understanding of the GBD?” YES/NO: 103 said yes, 0 said no, 103 respondents
• “What’s your overall rating of this course?” 4.6/5 average rating, 103 respondents

Organizational Impact

As a follow-up to this training I collaborated with researchers and data scientists to animate a data visualization of the spread of Omicron that received 2M impressions, 361K video views, and 81,065 engagements on Twitter alone. This was the most prolific media update in IHME’s history, indicating that we achieved the three primary organizational goals of the training:

• Educate journalists and policymakers on how to use and understand IHME’s data tools.
• Grow IHME’s scientific collaborator network.
• Reduce criticism from peers regarding a lack of transparency in IHME’s data modeling process.

Stakeholders’ Response

Starting the process with a minimum viable product allowed stakeholders to provide impactful feedback early in the design process, and led to their satisfaction with the deliverables.

As evidence of stakeholder satisfaction, the content was approved for translation to the six official languages of the UN: Arabic, Chinese, (English), French, Russian and Spanish, as well as Portuguese.

As of July 2022, I’m working with teams in Brazil and Mexico to localize content, translating to Portuguese (Br) and Spanish. This will allow learners to experience the content in their native language and help IHME reach a greater global audience.

Reflection

As an instructional/experience designer, I need to define learner and business needs, so that I can target solutions. This was a big change from my previous role creating sales/systems training at SEPHORA where all projects were tied to clear KPIs.

We eventually narrowed in on our target audience: scientific collaborators, with an understanding that other audiences would benefit from focused design. Although policymakers and journalists were <10% of our audience, they served a critical function in disseminating information and shaping public policy. They couldn’t be neglected. We found a balance by using plain language wherever possible and providing access to deeper information as needed.

Transparency of modeling techniques was a key concern for scientific collaborators. Although the content related to data modeling wasn’t easy to comprehend for policymakers and journalists, seeing that it was satisfying scientists’ needs provided a vote of confidence as indicated by press commentary.

I’m advocating for the use of conversions to the collaborator network as a leading indicator. If new learners join the network directly from the training, that’s another measurable outcome we can use to evaluate the success of the content.