Are p-values involved? If so, I think very much no.
-
-
-
So, it sounds like you're deciding what tests to run based on what your data look like. That's cool if you call it 'exploratory', but you need to run a new Exp if you want to test your new hypothesis. Or if you adjust your p-val for all the tests you [...]
1 reply 0 retweets 1 like -
[...] could have run, but didn't (and that's probably lots!). Andrew Gelman writes a lot about this as the 'garden of forking paths'/'researcher degrees of freedom', etc.
1 reply 0 retweets 1 like -
I can say more about this in the morning, if someone more qualified hasn't turned up by then!
1 reply 0 retweets 0 likes -
Replying to @TraversEoin @o_guest
The data are currently analyzed & the analyses are the same across the studies. We would keep the analyses the same if we make a different decision & it would necessarily be exploratory bc of combining scales. But thank you, I'll also try to consult with a stats person here.
2 replies 0 retweets 2 likes -
So, strictly speaking, reporting p-values for an exploratory analysis is a big no-no, unless you're adjusting for all of the other exploratory analyses might have tried. In practice though, people have done it for years, and you'll probably not get into trouble for doing it [...]
1 reply 0 retweets 0 likes -
[...] That is, it's technically wrong, but if you don't pretend you had decided a priori to combine the two measures and you're clear you're doing this post-hoc then it's not dishonest, and only crabit stats freaks will mind.
1 reply 0 retweets 0 likes -
[...] To summarise: In that case, go for it, but don't mention my name.
1 reply 0 retweets 0 likes -
Replying to @TraversEoin @o_guest
Lol thanks for elaborating. I will not mention you.
1 reply 0 retweets 2 likes
Out of context in my notifications that tweet was so weird. 
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
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.