Precisely Estimated


  • :belt_onion:

    While reading about some some people complaining about the way the Nobel Prize is awarded:

    https://www.theatlantic.com/science/archive/2017/10/the-absurdity-of-the-nobel-prizes-in-science/541863/?&single_page=true

    Another recent paper, which precisely estimated the mass of the elusive Higgs boson, has 5,154 authors.


  • area_can

    It could be a high precision estimate, e.g. 1-100 metric tons vs 1000-2000 Planck masses


  • ♿ (Parody)

    @el_heffe The science prizes can't be more absurd than the peace prize.



  • @boomzilla said in Precisely Estimated:

    @el_heffe The science prizes can't be more absurd than the peace prize.

    Clearly the duration of peace achieved is the culprit.

    "Congratulations, you won the election. Peace was achieved for 10e-27 seconds".



  • @bb36e said in Precisely Estimated:

    It could be a high precision estimate, e.g. 1-100 metric tons vs 1000-2000 Planck masses

    Neither of those is high precision.

    In scientific usage, "accuracy" and "precision" are related but not the same.

    Any measurement will have small variations if you repeat it. They might be smaller than the instrument's resolving power, but they are there. They will, in essence, cluster around some sort of "average" value.

    Accuracy measures how close that average is to the true value being measured.

    Precision measures the reciprocal size of the cluster (the more precise, the smaller the cluster).

    An instrument can be precise but inaccurate (e.g. it always reads 40 degrees high, plus or minus 0.01 degrees), or it can be imprecise but accurate (e.g. the measurements are scattered across a couple of hundred degrees, but by what amounts to blind coincidence, their average is less than 0.01 degree away from correct), but of course the most useful instruments are both precise and accurate.

    For an estimate, therefore, "highly precise" would imply that the uncertainty in the estimate is very small. So it might be "137 MeV +/-0.01 MeV", indicating an uncertainty of less than one part in 10,000.




  • area_can

    @steve_the_cynic isn't a range of 1-5mm more precise than 1-5cm, regardless of the accuracy of the measurements?


  • 🚽 Regular

    @Steve_The_Cynic I've always linked to this video for the difference: https://www.youtube.com/watch?v=hRAFPdDppzs



  • @boomzilla Or the so-called Economics prize.


  • ♿ (Parody)

    @slavdude said in Precisely Estimated:

    @boomzilla Or the so-called Economics prize.

    Which is not actually a Nobel Prize (maybe that's what you meant by so-called). But come on...Yasser Arafat?



  • @bb36e said in Precisely Estimated:

    @steve_the_cynic isn't a range of 1-5mm more precise than 1-5cm, regardless of the accuracy of the measurements?

    But the important thing usually is relative precision.

    A range of 1-5 mm on a measurement of micrometers is horrible precision. A range of 1-5 cm on something measured in km is great precision.

    And, on topic, the precision of those measurements is more like 1 part per million billion variance



  • @xaade said in Precisely Estimated:

    @boomzilla said in Precisely Estimated:

    @el_heffe The science prizes can't be more absurd than the peace prize.

    Clearly the duration of peace achieved is the culprit.

    "Congratulations, you won the election. Peace was achieved for 10e-27 seconds".

    https://www.youtube.com/embed/aRQLU3IwNYs?v=3&start=248&end=260



  • @boomzilla said in Precisely Estimated:

    But come on...Yasser Arafat?

    Maybe it's in relation to other PLO members?

    I mean, he was the only one that mentioned peace in his agreements, and it ended up splintering the PLO into multiple terrorist groups.


  • ♿ (Parody)

    @xaade You can rationalize it any way you want and I'm going to point and ridicule you.





  • @xaade Oh that clears it up.



  • @boomzilla said in Precisely Estimated:

    @xaade You can rationalize it any way you want and I'm going to point and ridicule you.

    I'm right there with you.

    I just wonder what caused people to think it was a great idea.



  • IMHO, the real wtf is the paper having 5000+ authors. How is that possible?


  • Considered Harmful

    Rubin and Franklin point to another longstanding issue with the Nobels. In as much as they propagate the myth of the lone genius, that lone genius is almost always white and male. Women have won just 12 of the 214 prizes in physiology or medicine, just 4 of the 175 prizes in chemistry, and just 2 of the 204 prizes in physics. The most recent female physics laureate, Maria Goeppert Mayer, won her prize 54 years ago. It’s not for lack of potential honorees, either. Rubin inarguably deserved one, as did Lise Meitner who contributed to the discovery of nuclear fission alongside laureate Otto Hahn. Between 1937 and 1965, Meitner was nominated 48 times by different people, and never won. “There are great things about the Nobel Prize but we should keep in mind that demographics of the winners reflect and amplify structural biases,” said astrophysicist Katie Mack on Twitter last year.

    There's the paragraph. I was looking for that. Simply because there are cool discoveries made by women does not mean that there are not better discoveries made by men at the same time; the article fails to adequately make its point here.



  • @zmaster said in Precisely Estimated:

    IMHO, the real wtf is the paper having 5000+ authors. How is that possible?

    In physics, the tradition is that everyone who worked on an experiment that produces a paper gets author credit, including theorists, beam operators, detector experts, data analysts, and a bunch of others. For experimental results at CERN, papers with several hundred authors are common.



  • @pie_flavor said in Precisely Estimated:

    Rubin and Franklin point to another longstanding issue with the Nobels. In as much as they propagate the myth of the lone genius, that lone genius is almost always white and male. Women have won just 12 of the 214 prizes in physiology or medicine, just 4 of the 175 prizes in chemistry, and just 2 of the 204 prizes in physics. The most recent female physics laureate, Maria Goeppert Mayer, won her prize 54 years ago. It’s not for lack of potential honorees, either. Rubin inarguably deserved one, as did Lise Meitner who contributed to the discovery of nuclear fission alongside laureate Otto Hahn. Between 1937 and 1965, Meitner was nominated 48 times by different people, and never won. “There are great things about the Nobel Prize but we should keep in mind that demographics of the winners reflect and amplify structural biases,” said astrophysicist Katie Mack on Twitter last year.

    There's the paragraph. I was looking for that. Simply because there are cool discoveries made by women does not mean that there are not better discoveries made by men at the same time; the article fails to adequately make its point here.

    For me, the rule of rewarding no more than three people distorts the proper credit, especially when the award goes to work that is the result of large collaborations. For the Higgs boson discovery, if you wanted to award the theorists, there were six scientists who could rightfully claim credit for the theoretical justification for the existence of the particle. Then again, the Nobel committee tends to favor experimental results, so the prize should have gone to CERN as a whole. If the Peace Prize can be given to organizations like Doctors Without Borders, then why can't the science prizes? This year's physics prize should go to LIGO, although this year's recipients make a lot more sense than the Higgs recipients.


  • Discourse touched me in a no-no place

    @mzh said in Precisely Estimated:

    the Nobel committee tends to favor experimental results

    Of course. Theorists have described a great many things over the years, most of which have had only a vague passing resemblance to reality.

    If the Peace Prize can be given to organizations like Doctors Without Borders, then why can't the science prizes?

    Different awarding institutions, different rules.



  • @zmaster said in Precisely Estimated:

    IMHO, the real wtf is the paper having 5000+ authors. How is that possible?

    Because in that field "author" doesn't mean "guy who sat down at a keyboard and typed stuff".

    The simple fact of the matter is proving their idea required this massive hardware apparatus, and virtually everybody who worked on building that is a co-author.



  • @mzh I agree.

    Imagine if a film award had the completely arbitrary restriction that the film could only include three actors. Or a video game award the restriction that the video game could only have three artists.

    It's ridiculous.



  • @blakeyrat said in Precisely Estimated:

    The simple fact of the matter is proving their idea required this massive hardware apparatus, and virtually everybody who worked on building that is a co-author.

    The problem I have with this is twofold: one, at which point in the chain of work do you stop? Is the janitor a co-author? What about the cook? The sub-contractors who built most of the stuff? I understand that you want to give credit to people who helped, and that might cause the author list to creep up from the usual 2-3, but how can they get to 5000+ and not stop and ask themselves if that is alright?

    (edit: a related issue is that it totally devaluates the fact of being an author. How can you differentiate between those who really did some work and those who just were cogs in the machine in such a list? Publication records is one of the metrics that is used currently to evaluate researchers, and such publications are making this metric useless for no good reason)

    The other problem is, did all these 5000+ people really read and comment on the paper? If not, they have their names on something that they did not actually review, and that isn't fair to them (when someone put me as co-author of a paper without mentioning it to me nor showing me the paper before publication, I was pissed), and they should be acknowledged in the paper, not co-authors. If they did really review it before publication, well, that's impossible to get 5000+ people to agree on a written text...



  • @remi said in Precisely Estimated:

    The other problem is, did all these 5000+ people really read and comment on the paper? If not, they have their names on something that they did not actually review

    Authors are not the same thing as reviewers. As @blakeyrat explained, the "authors" of a paper are the people who did the work described within the paper, one of whom will also have physically written the paper. That could be one person, or a team of 20, or a team of 5000, if the work is "build and operate a particle accelerator."

    Reviewers are completely separate things.



  • @remi said in Precisely Estimated:

    The problem I have with this is twofold: one, at which point in the chain of work do you stop? Is the janitor a co-author? What about the cook? The sub-contractors who built most of the stuff? I understand that you want to give credit to people who helped, and that might cause the author list to creep up from the usual 2-3, but how can they get to 5000+ and not stop and ask themselves if that is alright?

    If the people who came up with the concept in the first place are ok with 5,000 names there, then I don't see why the Nobel committee should be second-guessing their choice. That decision was already made and nailed-down before the Nobel committee got involved.

    @remi said in Precisely Estimated:

    (edit: a related issue is that it totally devaluates the fact of being an author. How can you differentiate between those who really did some work and those who just were cogs in the machine in such a list? Publication records is one of the metrics that is used currently to evaluate researchers, and such publications are making this metric useless for no good reason)

    Since basically the only metric academia has for success is "how many papers is your name on?" maybe this is someone being generous. Who fucking knows?

    Point is, the Nobel committee shouldn't be second-guessing why there's that many authors. The paper should be taken as a whole, or not considered at all.

    Look again, imagine Academy Awards. "We're giving this film Best Picture, but we're making it super-clear that its only because of the Director, Producer and Female Lead. Everybody else who worked on this film, fuck them." It's a dick move. Either you award the film as a whole, or not at all, there's no inbetweens.

    @remi said in Precisely Estimated:

    The other problem is, did all these 5000+ people really read and comment on the paper? If not, they have their names on something that they did not actually review, and that isn't fair to them (when someone put me as co-author of a paper without mentioning it to me nor showing me the paper before publication, I was pissed), and they should be acknowledged in the paper, not co-authors. If they did really review it before publication, well, that's impossible to get 5000+ people to agree on a written text...

    Again: that's the paper's sponsor's problem, not fucking yours. You don't get to tell them how to write their paper. Your nose is crammed WAY up in someone else's business. Yank it away.


  • 🚽 Regular

    @remi said in Precisely Estimated:

    @blakeyrat said in Precisely Estimated:

    The simple fact of the matter is proving their idea required this massive hardware apparatus, and virtually everybody who worked on building that is a co-author.

    The problem I have with this is twofold: one, at which point in the chain of work do you stop? Is the janitor a co-author? What about the cook? The sub-contractors who built most of the stuff? I understand that you want to give credit to people who helped, and that might cause the author list to creep up from the usual 2-3, but how can they get to 5000+ and not stop and ask themselves if that is alright?

    It isn't at all unreasonable to believe that 5,000 people were active in the design and assembly of LIGO or the LHC. And it is absolutely fair and correct to credit each of them for aiding in the discovery described in the paper. They are just as worthy of credit as one's lab assistant in a smaller scale experiment.

    @remi said in Precisely Estimated:

    (edit: a related issue is that it totally devaluates the fact of being an author. How can you differentiate between those who really did some work and those who just were cogs in the machine in such a list? Publication records is one of the metrics that is used currently to evaluate researchers, and such publications are making this metric useless for no good reason)

    Typically one has to discern their role in the experiment. If an intern screwed in some bolts to help construct the LHC while a PhD came up with the precise dimensions of a manifold in its design, that will definitely be factored as to whether one or the other is deserving of tenure at a university. AFAIK there is an order to the authorship of a paper, with the most significant contributor being first, and then going down in order of significance.

    @remi said in Precisely Estimated:

    The other problem is, did all these 5000+ people really read and comment on the paper? If not, they have their names on something that they did not actually review, and that isn't fair to them (when someone put me as co-author of a paper without mentioning it to me nor showing me the paper before publication, I was pissed), and they should be acknowledged in the paper, not co-authors. If they did really review it before publication, well, that's impossible to get 5000+ people to agree on a written text...

    The paper dispenses facts. The fact is someone was involved in the experiment, whether it was the design of the apparatus, interpreting its results, or being some kind of advisor or consultant, thus they are a co-author in the experiment and its results. If someone has some kind of objection to being listed as an author (e.g. they were unknowingly developing a Solaranite weapon ala Plan 9) then I don't know what their rights are in terms of whether they are entitled to have their names omitted from the paper, but because they were involved, by default they are listed as an author.



  • You could do it like movie credits: saying who did what.



  • @boomzilla said in Precisely Estimated:

    @slavdude said in Precisely Estimated:

    @boomzilla Or the so-called Economics prize.

    Which is not actually a Nobel Prize (maybe that's what you meant by so-called). But come on...Yasser Arafat?

    That, and the fact that it's used as a means to cement neoliberalism as economic orthodoxy.


  • ♿ (Parody)

    @slavdude said in Precisely Estimated:

    That, and the fact that it's used as a means to cement neoliberalism as economic orthodoxy.

    Seems like a feature to me. 🤷🏾♂



  • @blakeyrat said in Precisely Estimated:

    If the people who came up with the concept in the first place are ok with 5,000 names there, then I don't see why the Nobel committee should be second-guessing their choice. That decision was already made and nailed-down before the Nobel committee got involved.

    Yeah, if you're only talking about the Nobel committee, I agree. I was more discussing on the wider topic of having one paper with thousands of authors.



  • @gwowen said in Precisely Estimated:

    @remi said in Precisely Estimated:

    The other problem is, did all these 5000+ people really read and comment on the paper? If not, they have their names on something that they did not actually review

    Authors are not the same thing as reviewers.

    You are mis-reading what I wrote. There are "reviewers" in the strict sense of people who the editor ask to give their opinion on a paper that they were not involved in (and in that sense, it is stupid to even ask if reviewers have their names on a paper, the very principle of reviewing means that they don't!). And there is "reviewing something that has your name on it", as in "read and comment" about it, which is what I was mentioning.

    If you are OK with someone putting your name on something without letting you read the thing, well, someting is wrong.



  • @the_quiet_one said in Precisely Estimated:

    It isn't at all unreasonable to believe that 5,000 people were active in the design and assembly of LIGO or the LHC. And it is absolutely fair and correct to credit each of them for aiding in the discovery described in the paper. They are just as worthy of credit as one's lab assistant in a smaller scale experiment.

    Yeah but that doesn't answer a tiny bit my question:

    at which point in the chain of work do you stop? Is the janitor a co-author? What about the cook?

    Did they contribute less than the intern who screwed a few bolts?

    Again, I understand how you start by adding a couple people, and then a couple more, and again, and again. And it makes sense to explain reasonnably small authors lists, and there is no hard and strict rule about how many is too many. But clearly, 5000 is too many. That's just pointless.

    @remi said in Precisely Estimated:

    (edit: a related issue is that it totally devaluates the fact of being an author. How can you differentiate between those who really did some work and those who just were cogs in the machine in such a list? Publication records is one of the metrics that is used currently to evaluate researchers, and such publications are making this metric useless for no good reason)

    Typically one has to discern their role in the experiment. If an intern screwed in some bolts to help construct the LHC while a PhD came up with the precise dimensions of a manifold in its design, that will definitely be factored as to whether one or the other is deserving of tenure at a university.

    That is my point, how do you discern that? The typical method is to look at which papers you've co-authored as a first estimate. Here, you can entirely discard this paper even if it's ground-breaking because no one can know what your contribution was. Unless someone already knows your position in the internals of the project, but then your authorship doesn't matter since they know you already.

    So this kind of paper is pointless in terms of judging the scientific quality of someone, so it might as well be authored by "the CERN" or anything.

    AFAIK there is an order to the authorship of a paper, with the most significant contributor being first, and then going down in order of significance.

    Maybe. I think some papers are in alphabetical order (probably in some way to avoid ego problems because otherwise in a list of 5000 people you're bound to get some that are annoyed at being 4967th and not 4966th...).

    In any case, even if it's listed in the right order (and assuming you can actually define such an order!), there is no way they had an equal contribution so you are diluting the contribution of the first ones: as an outsider reading the paper, you can only assume that someone who's not last has done at least as much work at the person listed last, but you have no idea how much. So they at least screwed in a few bolts. Or maybe they invented the whole idea and are the next Einstein. Who knows? In other words, it is pointless.

    The paper dispenses facts.

    Have you ever been involved in writing scientific papers? Do you really think there is only one way to present things? And that there is no way that someone might want to add a part that someone else might not want to add, or present an argument in a different way, or simply word a phrase differently?

    If someone has some kind of objection to being listed as an author (e.g. they were unknowingly developing a Solaranite weapon ala Plan 9) then I don't know what their rights are in terms of whether they are entitled to have their names omitted from the paper, but because they were involved, by default they are listed as an author.

    I sure hope that they can have their name removed. Not that they should want it, but if someone has power to put my name as co-author of some text that I have not had the opportunity to review (as in, "give my approval") before it's published, I would be very pissed.

    Maybe the contribution of some people is under-represented because they were not liked by those who wrote the paper at the end (again, with so many people involved, this is bound to happen). Maybe their contribution is even mis-represented because the writers were not experts in that particular sub-field. The only way to avoid that is to have all co-authors read, comment and participate in the writing of the article (and in the end, have the option to withdraw their name if they feel the article doesn't represent what they think). This is the way it works everywhere (that I've seen or heard of). But with 5000+ people, this becomes impossible.



  • @remi said in Precisely Estimated:

    Again, I understand how you start by adding a couple people, and then a couple more, and again, and again. And it makes sense to explain reasonnably small authors lists, and there is no hard and strict rule about how many is too many. But clearly, 5000 is too many. That's just pointless.

    Why is 5000 clearly too many? What metric are you using to base that opinion on? Three zeroes is too many zeroes? The problem may be in your expectations. Both the document describing an experiment in a small lab with simple equipment and the document describing the operation of a 27 km long machine are both "papers" and follow the same publication guidelines, so you're trying to come up with a "reasonable absolute count of authors" that applies equally well to both.

    However, if an assistant in a simple experiment, who only manipulates simple lab equipment, gets to be co-author, why would the personnel required to design and run the experiment in the 27 km long machine, who probably have more training and had to put more effort and knowledge into it, not get to? Because you think 5000 is too big a number? Compared to what? Are you able to build and operate a massive particle accelerator using fewer people? Do you know of other massive particle accelerators that involved the contributions of 10 people?


  • Discourse touched me in a no-no place

    @xaade said in Precisely Estimated:

    @the_quiet_one

    0_1507141346782_da92f298-79a0-468f-9605-4806f83f6d70-image.png

    I prefer this one:

    0_1507218528471_e7416895-fe52-4681-a68e-60f7570ec30c-image.png



  • @kian said in Precisely Estimated:

    @remi said in Precisely Estimated:

    Again, I understand how you start by adding a couple people, and then a couple more, and again, and again. And it makes sense to explain reasonnably small authors lists, and there is no hard and strict rule about how many is too many. But clearly, 5000 is too many. That's just pointless.

    Why is 5000 clearly too many?

    Because this makes it impossible to judge the individual merit of the authors.

    you're trying to come up with a "reasonable absolute count of authors" that applies equally well to both.

    Yes, and I am trying to do so for one good reason, which is also the answer to the question "why do we list authors in a paper?"

    A paper has two (at least...) roles. One is to grow the scientific knowledge. In that sense, knowing the authors should not matter. If the paper is correctly written, all the information needed to reproduce the experiment (for experimental results) or all the equations (for theoretical ones) are in the paper, so science would be as much advanced without any name. So here, one name is as good as 5000, but so is no name at all.

    The other is to recognize the contribution of authors, both for simple moral reasons, and for practical reasons (such as estimating the "worth" of a researcher -- yes it's a flawed measure, but it still is not totally irrelevant and is the best we've found yet).

    And here, clearly, there is such a thing as "too many", which is what I've described in many examples in my previous posts (such as where do you stop, making it impossible to acknowledge the individual contributions, making it impossible for all authors to ensure their work is fairly represented etc.). And while, again, I don't have any clear and fast rule as to how many is too many, I am quite confident in saying that 5000 is too many.

    However, if an assistant in a simple experiment, who only manipulates simple lab equipment, gets to be co-author, why would the personnel required to design and run the experiment in the 27 km long machine, who probably have more training and had to put more effort and knowledge into it, not get to? Because you think 5000 is too big a number? Compared to what? Are you able to build and operate a massive particle accelerator using fewer people? Do you know of other massive particle accelerators that involved the contributions of 10 people?

    And why does the lab assistant get to be co-author and not the IT technician who set up the computer used to write the article, or the programmer who wrote the software to analyse the results? After all, he may have more training and put in more effort and knowledge than the lab assistant. Because you think setting up a computer is not related to the paper? Compared to what? He performed some work without which the paper could not have been written, and that the researcher would not have been able to do himself. Are you able to write that software yourself? Do you know of other data analysis software with comparable features involving 10 people only? See how your line of thinking can be reversed and applied to any case?

    You have to stop somewhere, otherwise you're going to list all the people in the world because they all somehow contributed to making it possible to write that paper. Is it arbitrary? Of course it is. Is it fair? Not always, sadly.

    But, for all the reasons I listed above, not doing it and listing 5000 authors means that being a co-author to that paper is entirely worthless.

    (note also in case it's not clear, your contribution can be acknowledged by other means than being an author -- I'm not saying these 5000 people did not contribute, I'm saying that having their name as authors is pointless)



  • @remi said in Precisely Estimated:

    authors

    I think you're having a problem more with the term "author" than the circumstance you're complaining about.

    If they were to call it "contributor" would you care so much?



  • @remi said in Precisely Estimated:

    I sure hope that they can have their name removed. Not that they should want it, but if someone has power to put my name as co-author of some text that I have not had the opportunity to review (as in, "give my approval") before it's published, I would be very pissed.

    But, the papers usually have findings, not opinions. You took part in the experiment, and this is what the experiment found. It's a fact.

    Now the experiment could be flawed or something, but I don't see how you'd be offended in someone writing up what they discovered.



  • @xaade said in Precisely Estimated:

    @remi said in Precisely Estimated:

    authors

    I think you're having a problem more with the term "author" than the circumstance you're complaining about.

    If they were to call it "contributor" would you care so much?

    I would still object if there is no separate list of authors. Read the rest of my post, I'm not objecting to the fact that these 5000 people contributed to the research, but to the fact that a huge list makes it impossible to split the contributions in any useful way, which is one of the role of the list of authors.

    If you're suggesting to have a small list of authors and a large list of contributors in a separate list, well, that's kind of what I was hinting to from the start when I spoke of "acknowledging" people (many papers have an "acknowledgement" section at the end which is used exactly for that -- to mention people who had a contribution, but are not authors, sometimes not because their contribution was small but because they were not involved in writing the paper or other reason).



  • @xaade said in Precisely Estimated:

    @remi said in Precisely Estimated:

    I sure hope that they can have their name removed. Not that they should want it, but if someone has power to put my name as co-author of some text that I have not had the opportunity to review (as in, "give my approval") before it's published, I would be very pissed.

    But, the papers usually have findings, not opinions. You took part in the experiment, and this is what the experiment found. It's a fact.

    Either you've never written a paper, or you're trolling.



  • @remi said in Precisely Estimated:

    that a huge list makes it impossible to split the contributions in any useful way, which is one of the role of the list of authors.

    That only matters if the "authors" contribute personal observations that exceed the contributions of the group. Given the paper isn't an opinion piece, it's completely irrelevant.

    If 3000 people worked on collating the data, and all that's in the paper is the data, then they're each an author.



  • @remi said in Precisely Estimated:

    @xaade said in Precisely Estimated:

    @remi said in Precisely Estimated:

    I sure hope that they can have their name removed. Not that they should want it, but if someone has power to put my name as co-author of some text that I have not had the opportunity to review (as in, "give my approval") before it's published, I would be very pissed.

    But, the papers usually have findings, not opinions. You took part in the experiment, and this is what the experiment found. It's a fact.

    Either you've never written a paper, or you're trolling.

    The conclusion part is very different from a vanilla opinion.

    Besides, the vast majority of papers I've read haven't listed out conclusions as definite things, but mostly possible things, and mostly describe correlations and not causation.



  • @xaade Then don't put any authors, or put a collective one such as "the CERN".



  • @remi said in Precisely Estimated:

    @xaade Then don't put any authors, or put a collective one such as "the CERN".

    You're getting stuck on the term "author".

    Should the 5000 participants come up with a team name, "Super Fire Alpha Squadron, GO". Would that make you feel better?



  • @xaade said in Precisely Estimated:

    @remi said in Precisely Estimated:

    @xaade said in Precisely Estimated:

    @remi said in Precisely Estimated:

    I sure hope that they can have their name removed. Not that they should want it, but if someone has power to put my name as co-author of some text that I have not had the opportunity to review (as in, "give my approval") before it's published, I would be very pissed.

    But, the papers usually have findings, not opinions. You took part in the experiment, and this is what the experiment found. It's a fact.

    Either you've never written a paper, or you're trolling.

    The conclusion part is very different from a vanilla opinion.

    Besides, the vast majority of papers I've read haven't listed out conclusions as definite things, but mostly possible things, and mostly describe correlations and not causation.

    OK, I have to assume that you've never written a paper then.

    There is a huge leeway in how you present results, whether you mention failed bits or not, how much emphasis you put on various parts and so on.



  • @xaade said in Precisely Estimated:

    @remi said in Precisely Estimated:

    @xaade Then don't put any authors, or put a collective one such as "the CERN".

    You're getting stuck on the term "author".

    Because this is what is seen in search engines, and what is used to compute citation indexes and so on. This is how it works. I'm not inventing stuff for the sake of the discussion.

    Should the 5000 participants come up with a team name, "Super Fire Alpha Squadron, GO". Would that make you feel better?

    Yes it would. Because then it's clear it's no longer the individuals but a group and nothing more, and it wouldn't pollute the "authors" information in various databases with useless information. But they already have a name, it's LIGO or CERN or something similar, no need to invent one to make you feel smart.



  • @remi said in Precisely Estimated:

    it wouldn't pollute the "authors" information in various databases with useless information

    Maybe everyone else understands that authors on a science paper means something different from authors on a fiction novel.

    For you, just label it "contributors" in your head, and be on the merry way.



  • @xaade Maybe. Maybe nobody looks at citation indexes and similar measures. Maybe cows are flying.


  • 🚽 Regular

    @remi said in Precisely Estimated:

    @the_quiet_one said in Precisely Estimated:

    It isn't at all unreasonable to believe that 5,000 people were active in the design and assembly of LIGO or the LHC. And it is absolutely fair and correct to credit each of them for aiding in the discovery described in the paper. They are just as worthy of credit as one's lab assistant in a smaller scale experiment.

    Yeah but that doesn't answer a tiny bit my question:

    at which point in the chain of work do you stop? Is the janitor a co-author? What about the cook?

    That's answered in the link I added to the post. Basically, if you're only involved in the maintenance of the equipment, it is unlikely to qualify you for inclusion. If you were involved in a more active role in the design of the equipment or experiment, then you are. So, no, cooks and janitors would not be included.

    @remi said in Precisely Estimated:

    Did they contribute less than the intern who screwed a few bolts?

    Again, I understand how you start by adding a couple people, and then a couple more, and again, and again. And it makes sense to explain reasonnably small authors lists, and there is no hard and strict rule about how many is too many. But clearly, 5000 is too many. That's just pointless.

    Thousands of people spent years of their lives designing and building one of the largest superstructures ever constructed for science. Are you saying we should arbitrarily omit who and who doesn't get credit for discoveries from that? I don't think you really realize just how big stuff like the LHC and LIGO are, and how many people had to use their scientific and engineering prowess to make it happen.

    @remi said in Precisely Estimated:

    (edit: a related issue is that it totally devaluates the fact of being an author. How can you differentiate between those who really did some work and those who just were cogs in the machine in such a list? Publication records is one of the metrics that is used currently to evaluate researchers, and such publications are making this metric useless for no good reason)

    Like I said, the most significant and important contributors are on the top of the list, and it goes down from there. Should Tom Hanks be offended that the right-most boy in the fourth row of the schoolbus be listed in the credits for Forrest Gump, too?

    Typically one has to discern their role in the experiment. If an intern screwed in some bolts to help construct the LHC while a PhD came up with the precise dimensions of a manifold in its design, that will definitely be factored as to whether one or the other is deserving of tenure at a university.

    That is my point, how do you discern that? The typical method is to look at which papers you've co-authored as a first estimate. Here, you can entirely discard this paper even if it's ground-breaking because no one can know what your contribution was. Unless someone already knows your position in the internals of the project, but then your authorship doesn't matter since they know you already.

    That's what recommendation letters are for. And CVs, which unless they're fabricated detail one's contribution to a project. If they do fabricate it, then it's a serious problem that will likely ruin their career, so they typically don't do that.

    AFAIK there is an order to the authorship of a paper, with the most significant contributor being first, and then going down in order of significance.

    Maybe. I think some papers are in alphabetical order (probably in some way to avoid ego problems because otherwise in a list of 5000 people you're bound to get some that are annoyed at being 4967th and not 4966th...).

    Let me ask you this: If someone applies for a job and says they worked for Microsoft, do you get all upset that you can't discern whether they were a mere "cog in the machine" or a principle architect? Or do you do what every friggen competent employer does and figure that out? You're treating this as if The Paper is the only sacred document that anyone can ever use. That there's nothing else in the world that can distinguish someone's role.

    In any case, even if it's listed in the right order (and assuming you can actually define such an order!), there is no way they had an equal contribution so you are diluting the contribution of the first ones: as an outsider reading the paper, you can only assume that someone who's not last has done at least as much work at the person listed last, but you have no idea how much. So they at least screwed in a few bolts. Or maybe they invented the whole idea and are the next Einstein. Who knows? In other words, it is pointless.

    No it's not. It's not a precise measurement. If you're 500th from the bottom versus 600th, the difference is negligible anyway. Hence what I said above about figuring out just what role someone had in an experiment.

    The paper dispenses facts.

    Have you ever been involved in writing scientific papers? Do you really think there is only one way to present things? And that there is no way that someone might want to add a part that someone else might not want to add, or present an argument in a different way, or simply word a phrase differently?

    In an experiment, you have different roles. One of these roles are writers. Saying that an intern should have equal say in how the document is written is as incorrect as saying an intern should have equal say in determining how many amps should be fed into one of the power couplings for the LHC. Again, using a movie analogy, almost none of the thousands of contributors to a movie has much say about how it's written or directed.

    Maybe the contribution of some people is under-represented because they were not liked by those who wrote the paper at the end (again, with so many people involved, this is bound to happen). Maybe their contribution is even mis-represented because the writers were not experts in that particular sub-field. The only way to avoid that is to have all co-authors read, comment and participate in the writing of the article (and in the end, have the option to withdraw their name if they feel the article doesn't represent what they think). This is the way it works everywhere (that I've seen or heard of). But with 5000+ people, this becomes impossible.

    First off, how is that impossible? These papers take years to write, get reviewed, and there are tons of drafts that need corrections. It's impossible to get a consensus, yes, but if your solution to the problem that things might get misrepresented if you have 5,000 contributors is to just have less contributors to be credited, then you're completely off base. It'd be even worse to have only the principle scientists write and review the paper and omit everyone else from not only the credits, but the review.

    At the very least, each contributor should be able to review their part of the paper. For instance, let's say a lowly contributor's sole participation was determining a margin of error for a specific part of the figures in the experiment. They provided that figure to the principle scientists and it was a critical piece of the experiment, and thus was included in the paper, which should be reviewed by the originator of the figure for accuracy. Then because that contributor's findings was an essential piece of the experiment, they are listed as part of the credit... because their figure is in the paper. If you omitted that person out from the credits, then it means you've just given that credit to someone else in the paper.


Log in to reply