Background

Google Play wanted to enhance context-aware app discovery for users. The challenge was identifying patterns in user behavior and improving recommendation algorithms.

Your Role

I conducted in-depth interviews and usability testing with over 20 participants, designed study protocols, and analyzed user feedback to inform design improvements.

Key Outcomes

• Enhanced user engagement with app discovery features.
• Optimized AI-driven recommendation algorithms based on user feedback.
• Presented actionable insights to stakeholders and the UX design team.

Research Methods

Usability Testing, Thematic Analysis, Interview Protocol Design

Visuals

Explore Google Play

Background

This project examines how people interpret data (i.e., graphical perception) by comparing different ways of visually representing it, such as static visuals versus motion-based techniques. The goal is to determine which approach or methods helps people understand numbers more effectively and accurately.

My Role

Designed and led the entire research process:

• Designed study protocols and recruited participants.
• Conducted two empirical studies (lab and Zoom-based).
• Analyzed data quantitatively.
• Published findings in IEEE Transactions on Visualization and Computer Graphics journal.
• Delivered a conference talk at IEEE VIS Conference (2024).

Key Findings & Outcome

• Proposed motion as a viable dimension for future visualization designs.
• Ranked motion-based encodings alongside traditional static encodings, providing a foundation for incorporating motion in duture data visualization design practices.
• Expansion motion ranked highly in user accuracy compared to static encodings.
• Non-synchronized motion offers better accuracy and faster response times than synchronized motion.

Research Methods

• Controlled Experiments (Designed and conducted two experiments to evaluate graphical perception of motion and static data encodings.)
• Quantitative Evaluation through System Log Analysis (Measured error rates and response times to assess perception accuracy and efficiency.)
• Within-Subjects Design (Ensured all participants experienced all conditions for consistency and control.)
• Statistical Analysis (Applied ANOVA tests and post-hoc comparisons to identify significant differences between conditions.)
• Bias Analysis (Investigated tendencies in over- or underestimation of encoded values.)

 

Read the full paper! [PDF]

Watch my presentation at IEEE VIS conference!

Review the user study!

Visuals