Increase bubble radius when BubbleView viewing time is short

Increase bubble radius when each image is shown only briefly. For example, on dense webpage-like images viewed for about 10 seconds, use a roughly 50-70 pixel bubble instead of a 30 pixel bubble so each click reveals enough context to sample the page.

A short timer limits how many clicks a viewer can make. If each click reveals too little, participants spend their time on mechanical exploration instead of reaching the most informative regions.

Mechanism: A larger bubble exposes more content per click. That reduces the number of clicks needed to cover dense layouts before the timer ends.

Evidence: On webpage images, BubbleView with 10 seconds of viewing performed worse with a 30-pixel bubble than with 50- or 70-pixel bubbles, and the paper reports that larger bubbles compensate when less time is available (Kim et al., 2017).

• User Goal: Approximate where viewers would look on a dense static visual.
• Task: Run a timed free-viewing BubbleView study.
• Data: Dense page-like layouts with many potential targets.
• Chart Setting: Blurred image with a fixed short viewing interval per image.
• Success Criterion: Participants can expose the important regions before time expires.

Break it when: The viewing interval is substantially longer, such as the 30-second condition tested in the paper. Why: Very large bubbles then expose too much at once and reduce selectivity, which lowered performance relative to smaller bubbles.

Sacrifice: You give up some spatial precision in the resulting importance map. Risk: A bubble that is too large can blur together nearby targets and reduce selectivity. Mitigation: Increase the bubble only enough to match the available time, or extend the time instead.

Mistake: Keeping a very small bubble in a short timed session. Why it fails: Participants spend their limited time opening adjacent small regions instead of reaching the important parts of the image.

Failure Sign: Participant traces show many adjacent clicks and key regions remain unopened when the timer ends. Quick Check: Inspect a few participant sessions and see whether they are still doing local exploration at timeout. Stronger Test: Pilot two bubble sizes and keep the one that yields a more stable hotspot map within the same short interval.

• Increase the bubble radius by one step and rerun a pilot.
• Keep the blur constant while adjusting the bubble so you isolate the time-radius tradeoff.
• If the bubble cannot grow without losing precision, lengthen the viewing interval.
• Recheck whether important regions are reached before the timer ends.