Thanks, this was interesting. I enjoyed your analysis and not having to read the paper myself. I happen to know that student mood tends to vary during the course of the semester, and there is a marked rise in mood after they pass. I would bet the missing data is from students who dropped the class.
I think the best way to test this is to make a cool mood testing app that people want to use, then give it permission to monitor usage of other apps. Have the person tell in words which media sites they use most often, and use that to monitor social media, because in my circle of knowledge only old folks do most of their surfing on facebook. It should monitor pinterest, tumblr, twitter, discord, etc. it should be distributed to non students. Just place it in the app store with all the permission forms preloaded after download,
Then again, I honestly don’t care what the study says. I find the important thing is limiting access to toxic sites, and not missing too much sleep that has the most affect on mood. Not the old “In my day we had real friendships” kind of moralizing that papers like to report.