In case it wasn’t obvious - from what I’ve read: you submit your paper to some journal (or conference). Peer review is arranged by the publisher/conference organizer, who gets a trusted group of people in the field to read the paper, ensure that it looks correct, not obvious flaws or misinterpretation of data, math errors, etc. If it passes, it gets published in the journal/presented and published in the conference proceedings. It’s not the peers’ job to verify the paper, simply to say “I can’t see any flaws in what they’ve done”)
there are occasional accusations or mutterings of poor reviews due to bias or rivalry, delays so the reviewer in the same field can pre-publish his similar paper, etc. Peer review is not intended to fully validate the findings, but to ensure the publication meets the standards of the field with no obvious flaws.
Another point is that some publications are valued (for hiring and promotions) based on the number of citations the paper receives. One thing I read joked that it was better to publish a controversial or slightly wrong paper, as so many people would cite that paper in attempting to refute it. (A publication references previous related work is relevant or applies to the topic at hand, in citations at the end and in footnotes)
Another joke, one of my in-laws did work in CERN, and mentioned that anyone who had a finger in the pie gets listed as author. So CERN papers, with dozens fo groups participating, will have a hundred or more “authors”. They are typically listed in alphabetical order, so having a surname starting with "Ab’ helps, since for brevit, the cite would be "by A. Abalone et al".
Trivially small these days; the Higgs discovery paper had 5,154 co-authors.
Even back in the day when such papers merely had dozens of authors, everybody knew that who happened to be first author alphabetically was meaningless. Yes, the first authors did have a certain celebrity amongst those reading the papers, but purely on the basis that it was obviously the spurious distinction.
Even in smaller groups, first author might or might not be the most prestigious position. In some disciplines, the guy who actually did most of the work is more likely to be the second author, or the last. I know that in relativity, at least, a two-author paper is likely to be a grad student (first) and his advisor (second), and decide for yourself which of those two actually has a better claim to authorship.
The grad student? Only slightly kidding. Who had the idea for a question and who did the work to create/prove the answer are both worthwhile questions.
The story as I learned it as an undergrad: Sharkovskii publishes his theorem in the Ukrainian Journal of Mathematics. No one outside the area reads it. Some years later, a prominent American researcher publishes a paper proving a very small special case of Sharkovskii’s Theorem (that period 3 points imply points of all periods), and its succinctness and applicability are such that he wins some award. Part of the moral was that it is much more interesting to have a small easily understandable result, and get published in leading journals, than to provide a huge amount of insight that leads to a very complicated, but thorough, statement of the nature of reality, that only gets published in a local journal.
Looking it up in Wikipedia, apparently the latter paper made a second claim beyond the first that was the special case of S’s Theorem, on a topic that wasn’t considered by Sharkovskii, namely that in the case of a period 3 point the map will exhibit properties they termed “chaos”. This latter point was probably why they got so much attention, as it started more research in chaos theory. But still, one might say that Sharkovskii proved far far more, but no one was interested in what seemed like just an ordering of integers giving you implications of periodic points.
Huh. With all of the bioinformatics papers I wrote or was involved with, the first author would be the student or research assistant who did most of the actual work (and was responsible for actually writing at least the draft of the paper), then the rest of the people who worked on the project/application/whatever would be listed, and the advisor / professor would always be the last author. So I got into the habit of basically only looking at the first and last author names…
Yeah, it really does depend. We used to give the person that did the work actually writing the paper the first spot. Usually this was the person that did the real research work anyway. If there was more than one additional author the remainder of the list was alphabetical. Papers that were true multi-author papers, we ran the author list in only alphabetical order. It is hard to come up with fair rules, and attempts to do fine grained ordering are guaranteed to end in tears.
When it comes to grad-student papers you can have just about the full spectrum of contribution. The brilliant student that the supervisor merely has to guide the work of, right through to the useless student where nearly every idea and every sentence has to be gone over multiple times with the student to beat some sense and readability into it (and the student.)
Right. It isn’t just a numbers game based on the number of publications you have. Hiring and tenure decisions are based on your citation index, the number of other articles that have cited you. (Citing yourself doesn’t count in the totals, and there is also an issue that citation indexes may not include book chapters and other non-journal publications.)
Publishing a controversial or wrong paper can be a good strategy to get cited a lot. But the champions are papers describing methods that get cited by everyone using that method. The all time champ is a 1951 paper by Lowry et al on a method of measuring proteins that now has more than 300,000 citations.
Another factor taken into account is the Journal Impact Factor of the journal where it was published. Publishing in a prestigious journal like Science or Nature is worth far more than publishing in a small local journal.
I’ve never experienced this. I tracked down my citation counts for my top papers and no one thought anything about them at all. One paper alone had more citations than most people in my department had for all their papers.
It was (officially) number of journal papers and what the outside letter writers thought. Unofficially it was always just total amount of grant money. (Which is against clearly written policy but …)
One compromise I’ve heard about (but apparently incredibly rare) is the candidate lists their 5 best papers and they are judged on those alone. People with a 100 crappy papers lose out, people with a few gems get a boost.
Quality over quantity, who ever thought that that might work?
I don’t actually sit on the committees that make these decisions around here, but people certainly talk about this all the time. Post-docs looking for a job are especially worried about it.
I’ve also seen researchers measured by a blended metric, usually called “impact factor” (though that can have other meanings, too): An author’s impact factor is the largest number n such that they have at least n papers which have each gotten at least n citations. So, for instance, if an author’s papers have gotten 10, 6, 3, 2, 1, 1, 1, 0 citations, then that author has an impact factor of 3, since they have 3 papers with 3 or more citations. If their next paper turns out to be a blockbuster and gets 73 citations, then their impact factor will still only be 3 (it can’t be 4 since their 4th-best paper still only has 3 citations).
This way, you won’t be unduly rewarded just because you happened to have one really interesting idea, if that’s all you ever do, and you likewise won’t be rewarded for cranking out a paper each week that each only get a single citation.
The h-index is pretty important in my field. Unsurprisingly, it’s a popular metric with scientific bureaucrats - boiling a scholar’s work down to a single number, how nice.
It’s less of a big deal in some fields - certainly, no self-respecting mathematician ever gave a fuck about their h-index, but in the chemical and life sciences it’s used a lot.
Despite its inherent fatuousness, it’s hard to deny it as a necessary condition of being an influential scientist, because how could anyone lay claim to that if no one’s reading and citing your work? It is not a sufficient condition, though - I know a few folk with h-indices of 40-50 (v respectable level in my area) who are fairly pedestrian. Hard workers who keep busy in thriving areas, but aren’t especially creative.
There are some exceptions where people have extremely low h-indices but have done amazing, Nobel prize level work. Either they died young, or more tragically, worked in industry. But you’re talking about a v small minority, esp these days.
Your Eros number is the number of connections it takes to get to Paul Erdős where a connection is having had sex with the person. If you had sex with him, your Eros number is one; if you had sex with someone who had sex with him, your Eros number is two; if you had sex with someone who had sex with someone who had sex with him, your Eros number is three; etc. Since he died a virgin, everyone’s number is infinite.
