IMRC IA Redesign
In this client-based course project, my team (3 members) were tasked with improving the International Migration Research Centre's (IMRC) information architecture to improve the site's user experience.
After conducting a content inventory, card sorting, and tree testing, we proposed several key changes to the navigational structure and information design of the IMRC with the intent of better supporting users' information seeking goals.
Role
Researcher
Lead Designer
Timeline
Sep - Dec 2023
Skills
Info Architecture
Tree Testing
Card Sorting
UI Design
Tools
Figma
Optimal Workshop
The purpose of the website is to establish an online presence where users can access scholarly articles written in affiliation to the IMRC, in addition to information about their mission, projects, and events. This gives audiences the opportunity to discover, interact with, and get involved with the organization.
Reasons Users Come to the IMRC Website:
1
To learn more about the IMRC's mission, researchers, and/or projects
2
To find scholarly articles pertaining to migration
3
To listen to the IMRC's podcast or watch their videos
4
To get involved with the organization
To gain a better context of the user behaviours of students, we conducted a review of several peer-reviewed research studies, finding that the main information-seeking behaviours used in this context were berry picking, foraging, and pearl growing. Thus, we knew that our changes to the website needed to fundamentally support these behaviours.
Information-seeking Behaviours We Need to Support:
Berry Picking
Where you do a query and get a long list of results
Then you click on a link, look around, and then come back to the results page
Then look again, and come back, look again, read more, etc.
You essentially go "picking and tasting berries" at each result to find the best ones
Foraging
Instead of nibbling berries at random, users optimize the clues (scent of information) in order to get to the juiciest food fast
Hunting for the best food (results) by looking across the plains and picking up on the clues that tell you where the best food is
Pearl Growing
When you find one result and it’s good but its not enough - you need more like it
"I found this, how do I get more like it?"
Just like how pearls grow by having more and more layers added
1
The site was not organized effectively
Some sections of the site felt hidden while other sections were outdated. For example, the News section of the site was unnecessarily difficult to navigate to, being hidden at the end of the spotlights section, while outdated sections of the website had entire top-level categories.
2
Labelling was inconsistent
The site often used various names for the same content, which can be misleading for users. An example was “News” and “Spotlights” being used interchangeably.
In other instances, labels did not match page titles. For example, clicking on “Projects Map” brought users to a page titled “IMRC Around the World”. While it may not seem like a large issue, inconsistent labels like these can make it difficult for users to find their way through the site. They might even question if they were brought to the right page.
There was also the issue of ambiguity for some labels. The label “Student Opportunities” sounded like a way for students to get involved; however, the content under it was actually about a scholarship award.
3
Navigation was unclear
Some of the site’s top-level categories had their own page and were thus clickable while others did not. This made the site’s navigation unclear.
Another problem was that breadcrumb trails were misleading, showing that users had clicked through more pages than they actually had.
What is a card sort?
Card sorting is a user research technique that helps us understand how users group content. These findings can the be used to inform information architecture.
How does it work?
Users are given cards with the titles of the site's content. Users arrange these cards into groups in a way that makes sense to them and then give those groupings a label.
Dendrogram (hierarchical representation of how users grouped content)
We learned that users primarily grouped content into 3 clusters:
1
Content about the IMRC itself
2
Content about the IMRC's research areas/focuses
3
Content about the IMRC's research endeavors (articles, podcasts, projects)
Based on how user’s grouped content, the labels they suggested, and other qualitative feedback, we proposed reorganizing content to fit under the following top-level navigation labels:
About
Our Focus
Research
Events
Get Involved
Contact Us
People
Pillars
Policy Points Publication
Projects Map
Displacements Podcast
Student Opportunities
What is a tree test?
Tree testing is a user research technique that tests a websites navigational labels and structure.
How does it work?
Users are given tasks and then have the ability navigate through the site's navigation labels to the point where they believe the information they are tasked to find would be situated. We can then examine whether users were successful and all the labels they clicked on.
Task 1
You were at a conference and heard about some of the IMRC’s work that they are doing in China. You want to know what other studies they are doing or have done in that area of the world. Where would you look to find this information?
Existing IMRC Navigation Structure PieTree (visualization of paths users took):
Proposed IMRC Navigation Structure PieTree (visualization of paths users took):
Key Results:
Task 1 Success increased from 70% to 100% using our proposed navigational structure
Task 2
You are pursuing a thesis on border issues and attended a seminar where the key speaker was a doctoral candidate studying with the IMRC. Their work has piqued your interest. Where would you look to find information on who they are?
Existing IMRC Navigation Structure PieTree (visualization of paths users took):
Proposed IMRC Navigation Structure PieTree (visualization of paths users took):
Key Results:
Task 2 Success only increased from 20% to 43% using our proposed navigational structure
Where did we go wrong?
From analyzing what labels users pressed on, we had a hunch that because we mentioned "seminar" in the task wording, participants looked for information about a seminar rather than the doctoral candidate.
To test this hunch, we tested Task 2 again with a revised task wording that did not mention a seminar.
Task 2 (Revised Task Wording)
While working on your thesis, you were told about a doctoral candidate studying with the IMRC. You want to find out more about what this person does, but you have forgotten their name. Where would you look to find information on who they are?
Existing IMRC Navigation Structure PieTree (visualization of paths users took):
Proposed IMRC Navigation Structure PieTree (visualization of paths users took):
Key Results:
Task 2 Success increased from 20% to 80% using our proposed navigational structure
This confirmed that the results in our previous test were negatively impacted by task wording, not the navigational structure
Our main improvements to the IMRC website:
A complete navigational restructure and content reorganization that improves navigability, matches user mental models, and maximizes information scent
Before
After
The site's new top-level categories in the navigation bar better match how user's expect information to be organized. This supports the site's navigability and helps users find what they're looking for faster.
Moving key information to the top of the page reduces "Iceberg Syndrome" where users can't tell what a site offers or is about upon immediate landing. In turn, this supports information forging and berry-picking by maximizing the information "clues" visible on the homepage
The incorporation of local navigation menus to support way-finding and user orientation
Before
After
Local navigation menus help users orient themselves in certain sections of the site while also telling them what other content is contained in this section.
In turn, this supports information foraging and berry-picking by providing information scent about the content offered on the site.
The addition of facetted search to make finding IMRC articles and projects hassle-free
Before
After
A search bar and search engine results page enables users to efficiently find articles matching their interests. Search facets also helps users narrow down results to the ones that best suit the user's interests.
This supports both pearl growing and information foraging as facets let users narrow down results according to their needs and results excerpts provide foragers with key information "clues".
A podcast page that supports discoverability
Before
After
Having brief excerpts open about the podcast episodes provides more information scent to be picked up by pearl growing and berry-picking information seekers. Users can quickly determine whether an episode piques their interest or not.
The addition of Get Involved, Events, and Contact pages to promote partnerships
After
After
Qualitative feedback from card sorting revealed participants felt these pages were missing on the site. As the IMRC is also focused on creating partnerships, the addition of these pages helps funnel users into ways they can get involved with the IMRC.
45%
Things I'd do next:
1
Redefine the existing design system or draft a new design system to match the proposed changes
2
Work on establishing a more consistent button design system across the entire website
3
Create an interactive prototype for usability testing
How I'd measure success:
1
Positive usability testing feedback
2
3
A/B Testing results confirming our changes yielded improvements