Here I am attempting to interpret notes from @manaswis so feel free to correct me on things.
The main idea is to use number of "breaks" that a user takes as the proxy for "diligence". Here, a break would be a stretch of time (maybe 30 seconds, maybe 60 seconds) where a user isn't interacting with the tool. Then we could compare average number of breaks per hour across user groups.
I'm honestly not sure I like the framing of taking breaks as a proxy for being less diligent. Breaks are good for you :+1:
@manaswis @jonfroehlich how do you feel about all of this?
Here I am attempting to interpret notes from @manaswis so feel free to correct me on things.
The main idea is to use number of "breaks" that a user takes as the proxy for "diligence". Here, a break would be a stretch of time (maybe 30 seconds, maybe 60 seconds) where a user isn't interacting with the tool. Then we could compare average number of breaks per hour across user groups.
I'm honestly not sure I like the framing of taking breaks as a proxy for being less diligent. Breaks are good for you :+1:
@manaswis @jonfroehlich how do you feel about all of this?