Some wonderful resources on how and why preregistration of statistical analyses is a fraught solution to reproducibility concerns.https://twitter.com/o_guest/status/1119609443153534982 …
-
-
Replying to @ctitusbrown
Or even why reproducibility itself might be a fraught way to drive science as a principle it and of itself... that's what Berna's talk/work is about, for example.
1 reply 1 retweet 3 likes -
Replying to @o_guest @ctitusbrown
If you can invent or deploy a new technology enabling experiments 10x larger (or more) than previously possible, you should do it. Then the previous result is a small special case. Not strict reproduction.
1 reply 0 retweets 0 likes -
Replying to @wait_sasha @ctitusbrown
That's kind of a different point to what I was making based on
@zerdeve's work. The gist is, if I understand correctly, that finding a true model of the world isn't made faster by optimizing for reproducibility. Check out the links in the thread for more info/her talk/paper.1 reply 1 retweet 4 likes -
I will watch the video -- thanks! Agreeing (I think?) that strict reproduciblity should not be the thing we optimize for.
2 replies 0 retweets 2 likes -
Replying to @wait_sasha @o_guest and
Just wondering if everyone in this discussion uses the same meaning for "reproducibility". Most misunderstandings are due to different definitions.
2 replies 0 retweets 2 likes -
Replying to @khinsen @wait_sasha and
Pretty sure there are various definitions in use but I don’t think what Olivia mentioned in her tweet can be attributed to those differences.
2 replies 0 retweets 2 likes -
Replying to @zerdeve @wait_sasha and
It's not only definitions, but also subjects. Reproducing an observation, a computation, and a finding/conclusion are three different things.
1 reply 0 retweets 0 likes -
And then it matters what the state of the art in the field is. Sometimes reproducibility means adopting established best practices, sometimes it would require going beyond, which might indeed not be a good investment.
2 replies 0 retweets 0 likes
Eh, I think you need to know about this work before tweeting TBH to have a constructive exchange here.
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.
