If you find a significant interaction, but sample size low, trustworthy or noise? I had a paper rejected that found profound but unexpected effect modification, it was not trusted/was dismissed by reviewers due to smaller sample size. @f2harrell
-
-
-
I have been involved in attempted replications of super strong interactions; the replications failed.
End of conversation
New conversation -
-
-
Take-home point: follow your professors on Twitter for a sneak preview of exam questions!
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Huh? Presumably this is only true for between-subjects designs, & under the (odd) assumption that interaction effect size is half main effect? In w/in subjects design, this just doesn't apply. Right?
-
eg: two spatial variables, both with two levels (Left,Right). keep L/R labels, then *interaction*=same>diff sides. Relabel 2nd variable columns as 'same/diff' side, then *main effect*=same>diff. *we* choose the labels!
End of conversation
New conversation -
-
-
I'm pretty annoyed I only memorized proofs in my statistics class instead of learning things like this.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I explain the issues here; it's a matter both of powering your smallest simple test of interest and estimating the interaction size from its shape:https://approachingblog.wordpress.com/2018/01/24/powering-your-interaction-2/ …
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
So how many countries do I need to have a cross-level interaction in my multilevel model? ;)
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Also important2note that this is NOT generally true4interactions w/ continuous modifiers, such as baseline BP or risk. Power can be excellent IF ppl stopped categorizing Xs when priori = continuous impact on effect size. https://www.ncbi.nlm.nih.gov/pubmed/16613605
- End of conversation
New conversation -
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.