Surely a correlation of 0.05 between social media use and depressive symptoms is evidence they're unrelated, if anything. It might be significant but N ~ 500k
-
Show this thread
-
Replying to @ed_berry
For sure any implication of causality is downright wrong. I agree most likely direction is depression -> social media /phone use. But for me explaining 5 (or 12) % of the variance of super complex disorder isn't a tiny effect?
1 reply 0 retweets 2 likes -
-
Replying to @o_guest @alex__morley
Those are the correlations though so wouldn't the r squared be 0.25% and 1.44%?
2 replies 0 retweets 4 likes -
Replying to @ed_berry @alex__morley
Should have clicked on your image.
So yeah, not sure why anybody takes this seriously as implying smart phones have anything to do with depression.1 reply 0 retweets 1 like -
The reason it's not OK to say 5% variance or whatever of a disorder is explained by smartphone or social media use is that without a theory of how this causes that or vice versa its just basically noise. If there's a theory I missed, then obviously that theory can be tested.
2 replies 3 retweets 6 likes -
Replying to @o_guest @alex__morley
Spoken like a true modeller
(that's a compliment)1 reply 0 retweets 1 like -
-
For the basic idea behind why correlation without theory is spurious think of these examples: http://www.tylervigen.com/spurious-correlations …
1 reply 2 retweets 3 likes -
I disagree that correlation without theory is spurious. Correlations can be a good entry point into discovering relationships, even very small ones. Of course, while theory isn't necessary for a .05 r to be meaningful, it's where you'd like to get to eventually.
1 reply 1 retweet 1 like
Olivia Guest | Ολίβια Γκεστ Retweeted Olivia Guest | Ολίβια Γκεστ
Olivia Guest | Ολίβια Γκεστ added,
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
