Great thread, with some insights that I hadn't thought about. Sharing code is good but it can lead others to use it uncritically. Include all relevant details that are needed to recreate your models in a paper. Then others can recreate your model from scratch to confirm results.https://twitter.com/SfPRocur/status/1102957227906949121 …
-
-
Yes, great thread. I wanted to hone in on "blind spot" b/c there's an "expert blind spot" when people in similar academic circles do have knowledge such as "common" assumptions even if they aren't in the paper, but those outside would not be able to replicate
-
so author & a part of the community will believe they have included sufficient relevant details, but it may not be sufficient for beginners/students, people not in the same field, etc. In short, "include all relevant details in spec" will always be subjective / depend on audience
-
so going back to initial topic: would open code help alleviate this bias? I think so, but it's a lot of work that isn't necessarily adequately rewarded
-
I don't agree open code in and of itself can fix this — I explain how and why in the thread and more depth here: http://dx.doi.org/10.1016/j.cogsys.2013.05.001 …
-
excellent, look forward to reading in depth
-
Hope it's useful!
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.