Share good ideas and conversation.   Login, Join Us, or Take a Tour!
comment by b_b
b_b  ·  335 days ago  ·  link  ·    ·  parent  ·  post: Social media and mental illness: "Nearly 63% of Instagram users report being miserable"

Meh, there's no control. No data from any of those charts show anything other than confirmation bias. Could be that miserable people are drawn to certain forms of social media.

veen  ·  335 days ago  ·  link  ·  

Maybe my title is a bit too brief - it is about whether someone is happy about the time spent on a particular app. I can think of two null hypotheses that the second graph alone seems to disprove:

- the amount of time users spend on social media platforms is unrelated to their reported happiness about using said app

- social media apps are not significantly different from phone calls / texts / emails

Those don't need a control group, right?

b_b  ·  335 days ago  ·  link  ·  

Both of your hypotheses, even if rejected, only point to correlated data. Correlation studies, especially those that rely on self-reported surveys, are very, very close to worthless, scientifically. FWIW, I happen to detest almost all social media, even the ones I use, so I don't find the conclusions of the study unbelievable--but it's just that the data they're using aren't what's making me believe it; it's my own experience.

You always need a control if you want to make a point that is believable. And not a historic control--a randomized, prospective control.