The quotes from Nature EiC Magdalena Skipper about whether journals should be checking for errors/data quality as part of peer review are quite surprising to me. https://www.wsj.com/science/whats-wrong-with-peer-review-e5d2d428?st=dhrnljoa74fujcv&reflink=desktopwebshare_permalink

What kind of signal is it to say that the data are so unimportant that they aren't worth checking during review? But big yay to the quote below. I'm hiring someone in my dept to do code and data check before submission. But we have lots of details to work out in protocol still.

This is all the more surprising given that in some fields (e.g., Economics) reviewers and readers routinely try to replicate analyses in papers before trusting the papers.

Sounds like Nature to me :)

Jeez

Then....then what is the job?

I agree with her that unpaid peer reviewers cannot be expected to trawl datasets. This means the journal needs to invest and pay for it.

What How about the technicians who check your plane for mechanical problems before you take it for a flight…they check the work and nobody feels adversarial to them.

Seems more in the realm of a Simpsons gag than a serious response, yikes!

I was once informally approached by one of the for-profit publishers about the possibility of starting a new journal as editor-in-chief. In my response, I requested money to hire freelance statisticians to review the statistical code, notation, and inferences. They said there wasn't money for that.

I don't find that comment surprising - publishers have made it clear that they don't think it's important to check for fraud