moin syed

@syeducation.bsky.social

Professor of Psychology, University of Minnesota. Sporadically writing stuff at http://getsyeducated.substack.com

@syeducation.bsky.social

Ira, all of this is fine, these are all reasonable positions, I don't know anything about this work. I made a narrow point about how you stated a malicious motive among the research team. I find that inappropriate absent any evidence, and unacceptable in the context of contradictory evidence.

@syeducation.bsky.social

Is it fair to mispresent what I wrote? I didn't say that the crisis "was never about replication rates." That would be absurd and I stated that the narrative was greatly accelerated by OSC 2015. The point is that the narrative existed prior to any rep rates, so is not **wholly** dependent on them.

In an otherwise interesting and useful analysis, Devezer & Buzbas unfortunately conclude their article with: (begin quote) Implications of our analysis extend to the broader narrative of replicability crisis. Declarations of crisis rest on the implicit inferential claim that observed replicability rates (such as the 36% reported in the Replicability Project: Psychology, Collaboration, 2015) are meaningfully lower than what a high-replicability literature would produce…. The crisis, if it exists, cannot be established by the methods used to declare it. (p. 28-29, italics in original). (end quote) This is a historically inaccurate claim, and one that I would otherwise ignore if not for the fact that such claims seem to be gaining in prominence. It has been jarring to see people attempt to rewrite a history that I both lived and is very well-documented. The declaration of a replication crisis, was not, in fact, based on the 2015 article. To be sure, publication of that article and the subsequent media coverage certainly increased attention to and use of the phrase (as Holzmeister et al. state), but by the time those results were known, “replication crisis” was already in common usage in psychology.
@syeducation.bsky.social

Thanks, agreed! This got confusing because upthread Matti (I think) accidentally referred to preprints rather than preregistration. What say you about preregistration on Zenodo?

@matti.vuorre.comReposted by @syeducation.bsky.social

Good talk. There's a lot of (over)thinking of this at the moment. Zenodo just works. (Until it doesn't when we and our bots ruin that as well.) Collaborate on GitHub and refer to zenodo.

@syeducation.bsky.social

Thanks, and very true. Many people value the "all in one place" criterion, and since that is not available they are weighing their options. I have been hearing a lot of interest in ResearchBox among those who value the "easy" criterion (e.g., they don't use GitHub)

Older posts