15 ms·
I'm surprised that there are still non double blind publications that count for tenure cases. This does explain some of the recent and embarrassing [lack of] r
by hedora 4y ago
I'm surprised that there are still non double blind publications that count for tenure cases.
This does explain some of the recent and embarrassing [lack of] retractions of Nature papers though.
- nequo 4y agoIt is nearly impossible to enforce double blind peer review because of the existence of online pre-prints (hosted on arXiv, SSRN, or just the authors' websites). A referee can copy and paste from the text into DuckDuckGo and find the authors.
- rgbrgb 4y agoIn this case an easy 50% solution seems better than nothing. Some people would just comply with the guideline, some would not think of your workaround and it would make peer reviews a little more equitable even if many broke the rule.
- hedora 4y agoI think you are overestimating how much reviewers care which grad student wrote which paper, or who their advisor is. Reviewers are generally overworked (for other reasons), and the main goal is to be constructive and somehow not sound like an idiot (in front of the other reviewers, who might be subject experts and definitely know who you are!) in any of the pile of reviews you need to write after reading a pile of papers. Also, reviewers declare conflicts ahead of paper assignments, which makes accidents less frequent.
- abdullahkhalids 4y agoBesides the sibling comment, in a lot of fields, people just know which group a paper is coming from even without doing any sort of searching or looking at preprints.
- bpodgursky 4y agoYeah it's often pretty obvious that a paper extends / is incremental on a previous body of work (maybe the same code, maybe same dataset) such that it could only have come from one lab.
- hedora 4y agoYeah, but good CS conferences have diverse reviewer pools, including ones from industry or different subdisciplines. Such reviewers often haven't heard of the lab, or paid attention to which grants are currently funded, the movement of students into postdoc positions at other labs, etc, etc.
- Blikkentrekker 4y agoI've seen many arguments against double-blind which were very unconvincing, and they were all from the people that already had a big name, the ones that decide how journals operate. In any case, this is the tip of the iceberg. The entire structure surrounding academic publication is absolutely ill-designed and few people in that world are even remotely interesting in safeguarding veracity. The overwhelming majority of peer reviewed scientific results is infotainment for other scientists with no party having any material stake in the veracity of it, which is almost no one bothers to replicate because accuracy is irrelevant so long it be interesting to read. The moment a company has bet a sizable stake on it's accuracy, then they suddenly check, and double check and have another party independently verify it because they do not want to loose money, obviously. But most science is nothing like that.
- runeblaze 4y agoJust to add more to the siblings. Research is inherently social. In a small field you can even guess the reviewers. Sometimes there is the "oh yeah that professor talks like that. It's definitely her writing me this review" moment, and reviews are supposed to be anonymous at least in most of the venues. Unfortunately the social dynamics and trust in some cases extend too far -- many sloppy research get published due to trust or favoritism. Double-blind makes this a bit better, but beyond this, at a certain point you will be changing what research is about from a social enterprise to something much more result oriented.
- cush 4y agoPeer reviews are blind, but in a lot of fields, once you've read enough papers it's obvious from subject and style who wrote the paper
- marlor 4y agoEven with double blind reviews, it’s often very clear who wrote the paper. Authors regularly cite work by themselves or their team, so a statement like “In a previous study[1] we established a relationship between X and Y” renders double blind pointless.