CoUX: Collaborative Visual Analysis of Think-Aloud Usability Test Videos for Digital Interfaces

Area of Science:

• Human-Computer Interaction
• Usability Engineering
• Visual Analytics

Background:

• Reviewing think-aloud usability test videos is time-consuming and requires UX professionals to analyze numerous user behavioral signals.
• Collaboration among UX professionals is essential for reducing bias and errors in usability evaluations, but it presents its own set of challenges.

Purpose of the Study:

• To introduce CoUX, a collaborative visual analytics tool designed to streamline the collective review of think-aloud usability test videos.
• To enhance the identification, annotation, and discussion of usability problems within an integrated environment.

Main Methods:

• CoUX visualizes problem-indicators derived from acoustic, textual, and visual features using machine learning to aid in usability problem discovery.
• A chatbox-like interface facilitates collaborative logging, commenting, and consolidation of identified usability problems.
• The tool's design was informed by a formative study with UX experts and literature review, followed by a user study with UX practitioners.

Main Results:

• User studies demonstrated that CoUX is effective in facilitating both independent problem identification and collaborative teamwork among UX evaluators.
• The study provided insights into how CoUX's features support both individual analysis and group collaboration.
• CoUX was found to be useful in making the collaborative analysis of think-aloud videos more efficient.

Conclusions:

• CoUX offers a valuable solution for improving the efficiency and effectiveness of collaborative usability test video analysis.
• The tool aids UX professionals in identifying and addressing usability issues more systematically.
• Further opportunities exist to enhance collaborative usability testing through advanced visual analytics and integrated communication features.