-
-
Notifications
You must be signed in to change notification settings - Fork 2
Description
Background
DICE currently tracks dwell time as the primary proxy of attention allocation, defined by content visibility within the viewport. A natural complement is to record which content intersects a horizontal reference line at a configurable position within the viewport. This "centerline content" measure operationalizes an assumption about visual attention: the element closest to the user's likely gaze region is the most probable candidate for active fixation, without requiring eye-tracking data.
Feature request
Implement a focal line exposure measure that tracks which post or content element is intersected by a horizontal line drawn at a researcher-specified vertical position within the viewport. The measure should record:
- Which element covers the focal line at each moment
- The cumulative duration each element spends at the focal line (analogous to dwell time)
The vertical position of the focal line (or area) should be a configurable parameter, expressed as a proportion of viewport height (e.g., 0.5 for the geometric center). This allows researchers to adapt the measure to their theoretical assumptions or platform-specific scrolling behavior.
Implementation notes
The viewport tracker already computes element positions relative to the viewport, so the focal line position reduces to viewport height × focal_line_position. For each tracked element, the check reduces to whether the element's top edge is above and its bottom edge is below this value. This requires minimal additional instrumentation beyond what the existing tracker captures.
Validation note
The focal line measure rests on the assumption that users' gaze tends toward a particular vertical screen region, which requires validation. Researchers who use this measure should not treat it as an established indicator of visual attention. They should justify their choice and interpret results with appropriate caution. That said, since implementation overhead is negligible, the feature is worth including so researchers can begin collecting the data and contribute to its eventual validation.