Google executives acknowledged this month they need to do a better job surfacing user-generated content after the recent Reddit blackouts.
Google executives acknowledged this month they need to do a better job surfacing user-generated content after the recent Reddit blackouts.
Don’t you also need to worry about your sample population being biased? You’d only be sampling people who sought out a dislike plugin, these people might be much more likely to dislike a video. Is there any way to account for that?
You’d have to have a separate cohort of non-plugin users & another with a sampling of both, I think. Run some regressions on those data and I think you’d be able to tease out any bias that exists.