UX Design UI Design

Introduction

About this Project

Ever struggled to find information on a website? We wanted to ensure UM Interactive Media’s site is intuitive and user-friendly. Through an online open card sort, we explored how users naturally categorize and label information, directly informing a more user-centered site structure. In this case study, I’ll walk through our research process, highlighting the crucial user insights we gained, and explain the actionable recommendations we developed to significantly improve usability, ensuring users effortlessly navigate and access the information they seek.

Research Goals

Taking notes from lectures or meetings is time-consuming. Many struggle to find specific moments in long recordings. Can AI solve this?

  1. Understand How Users Organize and Categorize Website Content:

    Why it matters: We want to learn how users naturally group and label information on the UM Interactive Media website to identify patterns in their mental models and improve navigation.

  2. Identify Pain Points in the Current Website Structure:

    Why it matters: Understanding where users struggle to find information will help us pinpoint areas where the site's information architecture needs improvement.

  3. Discover User Expectations for Website Navigation and Labeling:

    Why it matters: We want to determine the terminology and categorization methods that make the most sense to users, ensuring the website aligns with their expectations and enhances usability.

In my role as Lead Designer, I managed both the design process and project timelines, aligning the website's visual and functional aspects with study findings.

Team members

3

Project Length

2 weeks

My role

UX/UI Designer

My contributions extended to UX research, where I conducted participant interviews. These insights were instrumental in shaping design decisions and resolving key user pain points, all while adhering to project deadlines and milestones.

Methodology

Introduction & Consent

Before each session, participants received a detailed explanation of the study’s purpose and procedures. Informed consent was obtained, and explicit permission was granted to record the Zoom sessions for analysis.

Pre-Sort Questionnaire

Participants completed a 10-question demographic survey before the card sorting exercise. This provided contextual information to analyze potential correlations between participant backgrounds and sorting behavior.

Card Sorting Task

Participants used OptimalSort to categorize website-related items based on their intuitive navigation preferences. Moderators encouraged participants to verbalize their thought processes while minimizing interruptions. During the task, participants were asked to:

  • Name each group they created.
  • Suggest words users would expect to find on the homepage leading to those groups.
Post-Sort Activities

After completing the sorting exercise, participants:

  • Described their sorting strategy
  • Explored the UM Interactive Media website for five minutes.
  • Completed a post-sort questionnaire to provide additional insights into their experience.
Data Collection & Analysis
  • Category labels and group formations were recorded from OptimalSort.
  • Zoom session recordings were analyzed for qualitative insights.
  • Distance matrices were created for individual and combined participant data.
  • Tree diagrams were generated to visualize category relationships.
  • Participant comments were documented to identify common themes.
  • Findings were used to refine the website’s information architecture.
Test Environment & Moderator Role

All sessions were conducted remotely using Zoom and OptimalSort. Participants and moderators used personal computers with internet access. Each moderator facilitated four sessions, ensured the study was conducted properly, and documented participant insights.

Our pre-sort questionnaire helped us understand our participants' backgrounds, ensuring we captured a range of perspectives to inform our findings. We collected basic demographic details such as age, gender, and ethnicity to identify potential patterns in how different users organize information. Additionally, we gathered insights into their familiarity with technology and UM Interactive Media website to assess how their experience level might influence their sorting decisions. These responses provided valuable context for analyzing the card sort results, allowing us to draw deeper conclusions about user navigation preferences for the UM Interactive Media website.

Sampling Method:

Participants were recruited using convenience sampling, allowing us to efficiently gather a diverse range of perspectives and initial insights into user needs and preferences. However, future research with a larger and more representative sample will be necessary to validate these findings.

Participants:

We interviewed 12 participants, all existing or prospective college students over the age of 18. While the sample size was relatively small, it provided a valuable foundation for identifying key themes and informing the design of the UM Interactive Media website.

Card Sort Analysis

Our research revealed that participants were generally unlikely to recommend the current UM Interactive Media website, reflected in a Net Promoter Score (NPS) of -33.

Best Merge Method Dendogram

We considered a 40% agreement level as the threshold for identifying meaningful groupings. At this level, four distinct groups emerged, indicating that participants primarily organized content by function rather than audience:

  • Events and Sharing Spaces Information (yellow cluster) — This cluster shows strong agreement among participants, indicating that content related to events and shared spaces was naturally associated.
  • Application and FAQ (orange cluster) — The strong grouping at a low branch reflects that participants consistently connected application-related information and frequently asked questions.
  • Faculty/Accomplishments (blue cluster) — While less tightly grouped than the previous two, this cluster indicates a moderate consensus around content related to faculty profiles and student awards.
  • Career and Alumni Services (pink cluster) — Although this cluster formed at a higher branch, suggesting a weaker consensus, it still reflects a logical grouping based on career and alumni-related content.

Similar Matrix

Our analysis of the similarity matrix highlighted consistent patterns in how participants associated content:

  • Career-Related Groupings:

    Items like "Career services and internships for students" and "Posting jobs for students of the program" showed a high similarity score, indicating that participants consistently associated career services and job-related content as part of the same category.

  • Application-Related Groupings:

    Application requirements for international students" and "Information for international students applying to the program" were frequently paired, reflecting a strong user expectation that information related to international applications should be grouped together.

  • Campus and Facility Information:

    "Description of Meeting Spaces" and "Information about the building where classes are taught" had a high similarity score, suggesting that participants expected information about facilities and campus logistics to be presented together.

3D Clusters

Our analysis revealed six clear content clusters, reflecting user expectations:

  • External Career Source – Users linked "How to hire our students" and "Information for companies and organizations that partner with the program," indicating the need for a separate section for external partners.
  • Career Resources for Students/Alumni – Items like "Career services and internships for students" and "Posting jobs for students" were consistently grouped, suggesting a dedicated section for student career resources.
  • Overview of the Program – Strong connections between "Overview of the program", "Required courses", and "Program concentrations" highlight the need for a clear program overview section.
  • Admission & Application – Clustering of "Online application," "Application requirements," and "Program tuition" supports consolidating all admission-related content.
  • Research Spaces – Items like "Description of research facilities" and "Meeting spaces" were grouped together, suggesting users expect facility details to be unified.
  • Newsletter & Social Media – Grouping of "Newsletter sign-up" and "How to follow the program" reflects the need to centralize communication resources.

Analysis

The findings from our card sorting exercise provided valuable insights that directly influenced our design recommendations. Due to project limitations, we were unable to develop a fully interactive website. However, leveraging the analysis, we were able to identify key secondary navigation patterns and make informed recommendations for a future build.

  • Prioritize Career: Maintain prominent "Career" section based on user importance.
  • Enhance Admissions: Continue dedicated "Admissions" with detailed info, especially for international students (Miami relevance).
  • Clarify Research/Facilities: Standardize terminology, group research under "Research" for better navigation.
  • Maintain FAQs: Keep separate FAQs for Career and Admissions for easy access.
  • Support Communication: Retain "Newsletter" and social media links for engagement.
  • Organize Programs: Keep the current clear program breakdown.
  • Expand Student Life: Develop a comprehensive "Student On-Campus Life" page.

Secondary Navigation

Final Prototype

View Prototype

Reflection

Insights from Card Sorting Study

Conducting the study reshaped my approach to UX design by revealing how users naturally structure information. It pushed me to rely on data rather than assumptions.

Impact on Navigation & Structure

Seeing users group related concepts led me to create a "Career & Internships" section, improving navigation and accessibility for career resources.

Key Takeaways on Mental Models

This experience reinforced the need to design in line with users’ mental models, ensuring information is presented intuitively for better usability and satisfaction.

P rev Note taking with AI & Audio
Say hi