what do i doooo with my e.g. r=.09 correlations at 480k sample size? it's definitely a Real Correlation, especially when there's a cluster of similar correlations, and they're mildly interesting, but a little *too* mild if u know what im sayin
Conversation
Replying to
You are so deep into self selection territory this is not as useful as random by any means
2
19
Show replies
Replying to
how are you determining whether the correlation is real? are you compensating for multiple testing?
1
13
Show replies
Show replies
Replying to
Post on twitter or similar, that you found X correlation with an r of .09, but it isn't interesting enough for you to explore any further.
2
Show replies
Replying to
Hmmmm. No mention of F-stats in the comments. t-stat applies to the coefficient; F-stat to the R2. You can have a small but significant correlation and vice-versa.
1
