As long as everybody is 100% clear about the model not having undergone a test in phase B, and about any inferential statistics in phase B being problematic, I am completely on board with that idea.
-
-
Replying to @VandekerckhoveJ @EJWagenmakers and
Why not use existing modelling evaluation techniques like e.g., cross validation, etc.?
1 reply 0 retweets 2 likes -
-
Replying to @VandekerckhoveJ @o_guest and
Indeed, but for ols one can get loo from the
1 reply 0 retweets 0 likes -
Replying to @wdonald_1985 @VandekerckhoveJ and
hat matrix. Bayes has apprx. loo
1 reply 0 retweets 0 likes
Replying to @wdonald_1985 @VandekerckhoveJ and
Sorry. I mean I don't follow how to prereg a model. (I understand the cv data requirements aspect.)
11:42 PM - 22 Nov 2018
0 replies
0 retweets
0 likes
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.