Classification of Gallery View Scanning
A practice of reading facial expressions across a grid of video thumbnails has developed without any consensus on what is being measured.
Meeting Minutes: Gallery View Scanning Review April 2022
Present
Fourteen participants. Grid dimensions: four by four, minus two participants on audio only, minus one screen share that replaced a face with a spreadsheet. Effective observable faces: eleven.
Item 1: The Practice
It was noted that participants in video calls routinely scan the gallery view grid during presentations and discussions. The scan occurs approximately every thirty to ninety seconds, depending on speaker and content.
The purpose of the scan was identified as: checking whether others are engaged, checking whether the speaker is being received well, checking whether anyone appears to disagree, and checking whether anyone has left the call.
These are four distinct activities performed in a single pass lasting approximately two seconds.
Item 2: Observable Behaviors Under Review
The following behaviors were documented across eleven observable faces during this meeting:
Looking slightly left of center: Seven participants. This is consistent with looking at a second monitor, reading a transcript, or looking at their own face.
Nodding: Three participants. The nodding is regular and has continued for the duration of the meeting without obvious relationship to the content.
Neutral expression: Eleven participants at various intervals.
Item 3: Interpretive Disagreement
The scanning practice assumes that faces are legible and that the information conveyed is useful. Neither assumption was confirmed during this meeting.
The faces are present. The faces are small. What the faces mean is a matter of ongoing disagreement.
Item 4: The Camera Problem
Camera placement on laptops positions the lens above the screen. Participants looking at the face of the person speaking appear to be looking down. Participants looking at the camera appear to be looking at the ceiling.
This means that attentiveness and inattentiveness produce similar visual profiles, which are then scanned every thirty to ninety seconds by eleven other people.
Item 5: No Action Required
The committee did not identify a proposed action. The practice continues.
Minutes recorded.