Pre-reg is just an easy way to make your specification public and safeguard against post-hoc changes.
-
-
Replying to @annemscheel @Darren_Rhodes
it's part of a school of thought just like anyth is. There's no single scientific method http://plato.stanford.edu/entries/scientific-method/ …
1 reply 0 retweets 0 likes -
also a model's specification is not known until after you have done all the work!
1 reply 0 retweets 0 likes -
Curious to know whether you would consider the work you do "hypothesis testing" in a deductive sense.
1 reply 0 retweets 2 likes -
I do modelling that fits into that and modelling that doesn't
1 reply 0 retweets 0 likes -
Do you use p values to draw inferences in your work?
1 reply 0 retweets 0 likes -
I have not done a frequentist test on a model as far as I can recall
1 reply 0 retweets 0 likes -
Replying to @o_guest @chrisdc77 and
Bayesian statistics are only immune to overfitting you have the correct model and priors
2 replies 0 retweets 0 likes -
Replying to @talyarkoni @chrisdc77 and
also just to be clear I'm not a Bayesian myself, but not anti Bayesian either.
1 reply 0 retweets 0 likes -
Replying to @o_guest @talyarkoni and
I don't do anyth Bayesian. I obviously can't rule it out for future work, but I haven't.
2 replies 0 retweets 0 likes
I've mainly done connectionst. Not ideological connectionst though. If anyth mildly anti.
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.