Card Sorting Study:

Interactive Media Website 

IA Redesign

Role (Team-Based Work)
Team Leader
Led the card sorting study.
UX Designer
Contributed in data analysis
and IA redesign.
One Week
Open Card Sorting
Affinity Diagram
Microsoft Excel

Well-organized information architecture can help information to be found easily and used effectively while poorly-organized information architecture can cause user failure. For the Interactive Media website, which provides diverse information about the program, it is critical that the website has its content well organized. To evaluate the information architecture of the website and to recommend a new IA design, we conducted an open card sorting on the Interactive Media website. 

Research Process
Data Collection
Data Analysis
Product Description
Screen Shot 2019-02-24 at 8.20.27 AM.jpg

The Interactive Media Program Website

The Interactive Media Website is a website that introduces the Interactive Media Program at the University of Miami to visitors. There are four main audiences it targets: prospective students, current students, alumni, and donors. The website serves these four general audiences in their pursuit to apply to the School of Communication’s Interactive Media program, find jobs and internships, learn about the facilities, find opportunities as alumni, or even make a donation to this program.


Current IA Issues

In order to determine the current information hierarchy problems of the website, I created a sitemap to identify its structure.

Interactive Media Before.png

Issues of the current IA

Four information architecture issues were found:

ISSUE 1: Inconsistent sidebar hierarchies across pages.

ISSUE 2: Conflicting content titles and sidebar menu labels.

ISSUE 3: Repeating irrelevant content across pages.

ISSUE 4: Missing category landing pages.

Uncovering Users’ Mental Model of Information Structure

To create an information structure that matches users’ expectations, an open card sorting was conducted on 48 participants through OptimalSort. The content on the website was described separately on 48 cards. The procedure of the experiment was as follows:

•   A participant first arranged the cards into groups.
•   Once finishing sorting cards, the participant was asked to name each group.
•   The participant was asked to explain their grouping strategies after submitting the result to OptimalSort.
•   The participant was then asked to create a hierarchy of these groups.
•   Prioritizing strategy and additional comments of the participant were later collected.











About Us




Contact Us






50% Male
50% Female
Avg Age: 25.81
CardSort-demographic viz.png
How Did Participants Relate the Content?

A combined analysis of the similarity matrix, the dendrogram, the grouping strategies, the prioritizing strategies of the participants as well as the labels used for all groups of cards reveals their understanding of the content.

Group 9.png
Results of the similarity matrix were used to identify possible groups.
Similarity Matrix
Similarity Matrix Addomg pulling Groups
Group 8.png
Group 7.png
Subcategories were identified based on the relevant level of each item using results of the actual agreement method dendrogram.
Actual Agreement Method Dendrogram
Screen Shot 2019-10-20 at 7.49.16 PM.png
By using an affinity diagram, we concluded there were 5 grouping strategies adopted by our participants. These 5 strategies helped us to finalize 8 categories.

5 Grouping Strategies:

  • Numbers of groups

  • Audience-based categorization

  • Relevant at the time

  • From general information to specific information

  • Keyword-based categorization

Affinity Diagram for Grouping Strategies
Grouping Strategies.png
Group 6.png
Group 6 Copy.png
 Key Findings
Results of prioritization along with participants’ prioritizing strategies helped identify the orders of the 8 categories as well as the items within each category.

2 Prioritizing Strategies:

  • General info prior to specific information

  • In the order of relevance to a type of audience's goal

Affinity Diagram for Prioritizing Strategies
Prioritizing Strategires.png
We also analyzed participants’ preferred terminologies for the label of each category
Analysis of Preferred Terms

Direct access to information & audience-based.

Information for different audiences should be separated. Direct access to all the information for one type of audience always helps.

Keywords in labels matter

People tend to find information based on keywords, so consistent terminology should be used when describing related content.

Prospective students would like to first learn about the program generally before they look into the details and decide whether or not to apply.

As the major user group for the website, prospective students prefer information that can give them a better general idea of the program before looking into other details. Therefore, information ranging from what students can learn from the program to its academic level, as well as career opportunities, facilities and resources should be well-organized and easily accessible.

Improved Information Architecture

Based on the findings of the open card sorting, we provided an improved information architecture of the website:

Below are the redesign recommendations:

Add a landing page to each section.

A landing page containing general information, such as “Academics description,” ”How we can help you launch your job,” describes the scope of this menu item and gives users a better understanding of what this section is about. This can help increase users’ confidence to use the website for information.

We also provide direct access to the most important messages on the landing page, for example, application deadline and application prerequisites. In this way, users do not have to go to each sub-topic page to look for the messages.

Keep the sub-topic menu structure of each section consistent.

The current site has an inconsistent sub-topic menu structure across sections. In some sections, the menus consist of titles of all the sub-topics while in other sections, like FAQ and Contact Us, the menus are a copy of the main navigation items.

We recommend that each section have a landing page first and then a list of menus for its sub-topics. For those who do not have sub-topics, we recommend putting all the content on the landing page and getting rid of the sub-topic menus.

Sort the groups according to users' mental model & use related keywords as menu labels.

Participants' grouping and prioritizing strategies revealed the pattern of orders in which they looked for information: Application -> Academics -> Career. Therefore, we broke down the previous About Us main navigation item and reorganized its content into related sections. All information regarding career and the school’s current partnerships was grouped into the ”Career & Collaboration” section.

We relabeled ”Admissions” to ”Application” as the latter was much more preferred than the former.