Web pages: What can you see in a single fixation?

Area of Science:

• Cognitive Psychology
• Human-Computer Interaction
• Computer Vision

Background:

• Human vision efficiently processes natural scenes, extracting semantic category, spatial layout, and object identities in a single fixation.
• This rapid scene understanding is crucial for navigation, threat detection, and guiding further visual exploration.

Purpose of the Study:

• To investigate whether humans can extract gist information from artificial stimuli like web pages at a glance.
• To determine if website type and element locations are discernible with brief visual exposure.

Main Methods:

• Participants viewed web pages for only 120 milliseconds, with fixations centered.
• Performance was assessed on classifying page categories and localizing specific elements like menu bars and advertisements.

Main Results:

• Observers could classify web pages into one of ten categories significantly above chance levels.
• Participants demonstrated an ability to understand spatial layout, accurately locating menu bars and detecting ads.
• Text readability at a glance partially supported this gist comprehension.

Conclusions:

• Web pages, despite being artificial, can be understood at a glance, similar to natural scenes.
• Findings suggest parallels between web page gist and scene gist, with implications for understanding visual processing and designing user interfaces.