Are there never mistakes in peer reviewed science? No, there are
The question isn’t “are there zero mistakes”, the question is, “is peer reviewed research generally of higher quality than blogposts?”. To which the answer is obviously yes (at least in my opinion), although the peer review process is cumbersome and slow, and so will have less output and cover less area.
When there are both peer reviewed research and blogposts on a subject matter, I think the peer-reviewed research will be of higher quality and more correct a vast majority of the time.
Upvotes on an internet forum are not a good replacement for peer review. I’m surprised I even have to argue for this, but here goes:
the vast majority of people upvoting/downvoting are not experts in the topic of the blog post.
The vast majority of upvoting/downvoting occurs before a blogpost has been thoroughly checked for accuracy. If theres a serious mistake in a blogpost, and it’s not caught right away, almost no-one will see it.
Upvoting/downvoting is mostly a response to the percieved effort of a post and on whether they personally agree with it.
Yes, peer review is flawed, but the response isn’t to revert to blogposts, it’s to build a better system.
I think that peer review is so poor that probably just the forum alone produces work that is less in need of replication. I guess that’s not really about the system.
And yes, we should build a better system, but still. Peer review vs upvotes on journal sites, I would pick the latter.
The question isn’t “are there zero mistakes”, the question is, “is peer reviewed research generally of higher quality than blogposts?”. To which the answer is obviously yes (at least in my opinion), although the peer review process is cumbersome and slow, and so will have less output and cover less area.
When there are both peer reviewed research and blogposts on a subject matter, I think the peer-reviewed research will be of higher quality and more correct a vast majority of the time.
Compared to EA blog posts weighted by karma? The Answer is not obviously yes in my opinion. I think we’ll fare better in the replication crisis.
Upvotes on an internet forum are not a good replacement for peer review. I’m surprised I even have to argue for this, but here goes:
the vast majority of people upvoting/downvoting are not experts in the topic of the blog post.
The vast majority of upvoting/downvoting occurs before a blogpost has been thoroughly checked for accuracy. If theres a serious mistake in a blogpost, and it’s not caught right away, almost no-one will see it.
Upvoting/downvoting is mostly a response to the percieved effort of a post and on whether they personally agree with it.
Yes, peer review is flawed, but the response isn’t to revert to blogposts, it’s to build a better system.
And yet argue it you shall.
I think that peer review is so poor that probably just the forum alone produces work that is less in need of replication. I guess that’s not really about the system.
And yes, we should build a better system, but still. Peer review vs upvotes on journal sites, I would pick the latter.
Maybe we could discuss it in the comments of https://forum.effectivealtruism.org/topics/peer-review