Slashdot is powered by your submissions, so send in your scoop

 


Forgot your password?
Close
typodupeerror
×
Science

Russian Journals Retract More Than 800 Papers After 'Bombshell' Investigation (sciencemag.org) 42

Posted by msmash from the how-about-that dept.
sciencehabit writes: Academic journals in Russia are retracting more than 800 papers following a probe into unethical publication practices by a commission appointed by the Russian Academy of Sciences (RAS). The moves come in the wake of several other queries suggesting the vast Russian scientific literature is riddled with plagiarism, self-plagiarism, and so-called gift authorship, in which academics become a co-author without having contributed any work. The RAS commission's preliminary report documenting the problems and journals' responses to them is "a bombshell," says Gerson Sher, a former staffer at the U.S. National Science Foundation and the author of a recent book on U.S.-Russia science cooperation. The report, released yesterday, "will reinforce the suspicions and fears of many -- that their country is not going down the right path in science and that it's damaging its own reputation," says Sher, who applauds RAS for commissioning the investigation.

Russia's roughly 6000 academic journals, the vast majority published in Russian, are popular among the country's academics. A 2019 study found that Russian authors publish far more in domestic journals than, for instance, their counterparts in Poland, Germany, or Indonesia. But standards are often low. In March 2018, for instance, Dissernet, a network aimed at cleaning up the Russian literature, identified more than 4000 cases of plagiarism and questionable authorship among 150,000 papers in about 1500 journals. And Russian authors frequently republish their own work, says Yury Chekhovich, CEO of Antiplagiat, a plagiarism detection company.
This discussion has been archived. No new comments can be posted.

Russian Journals Retract More Than 800 Papers After 'Bombshell' Investigation More

Russian Journals Retract More Than 800 Papers After 'Bombshell' Investigation

Comments Filter:

  • Duh! (Score:5, Funny)

    by methano ( 519830 ) on Wednesday January 08, 2020 @12:49PM (#59599418)
    The scientific publishing machine is also pretty bad in the US and China. I've been watching it up close for about 40 years. The bottom line is that you get what you reward.

    There was a recent report that about 80% of the english language scientific biomedical literature was not reproducible. I'd cite the report but there's only a 20% chance it's right.

    • The scientific publishing machine is also pretty bad in the US and China. I've been watching it up close for about 40 years. The bottom line is that you get what you reward. There was a recent report that about 80% of the english language scientific biomedical literature was not reproducible. I'd cite the report but there's only a 20% chance it's right.

      Definitely starting to get a dog-bites-man vibe from these kinds of stories.

    • Replications don't get published (Score:5, Insightful)

      by shanen ( 462549 ) on Wednesday January 08, 2020 @02:09PM (#59599754) Homepage Journal

      I'd give you the funny mod if I ever had a mod point to give.

      However I recently read a very interesting analysis of the problem in terms of selection bias. Replication studies are important, but easy to reject. Basically "Nothing to see here." Results that seem new and interesting are much more likely to be accepted for publication even if they are shaky. This was described in the context of trying to write surveys of the literature that correct for the bias.

      Having said that, the problems in Russia seem to be more systematic and have deeper roots. Some of the distortions (especially in biology) had ideological sources, and some were driven by strange economic limitations (as in computer science).

      • Re: (Score:3)

        by ceoyoyo ( 59147 )

        People have this idea that a published study is supposed to be "right." It's not. It's a report of something that's potentially interesting. It *needs* to be replicated.

        The fact that much research is not replicable is more a warning that we're not doing enough replication work.

        The problems in the actual story are different. Despite the anti-Russia slant, they don't really sound any worse than what's been found in other journals. There's *one* Japanese guy, publishing primarily in big international journals

        • I agree in that the ethical issues should be accuracy and honesty about what was done and observed, NOT finality. There should be nothing wrong with a paper that says, "here's what we did and here's what we saw - these patterns were observed within our sample" - just so long as it's not misrepresented as more than it is. Re-publishing one's own work is also not unethical so long as it is not represented as something else.

          (Of course, just because your work avoids ethical violations doesn't mean it is im

      • Re:Replications don't get published (Score:4, Insightful)

        by larryjoe ( 135075 ) on Wednesday January 08, 2020 @03:31PM (#59600086)

        Replication studies are important, but easy to reject. Basically "Nothing to see here." Results that seem new and interesting are much more likely to be accepted for publication even if they are shaky. This was described in the context of trying to write surveys of the literature that correct for the bias.

        To further emphasize this point, in certain research areas (e.g., computer architecture and many related areas), replications studies are never accepted as-is unless the replication data happens to be included as an ancillary part of some other main, novel contribution. Reviewers rate papers based on novelty, and replication studies get zero credit for paper acceptance. That is, it's not just that reviewers are individually rejecting replication studies; it's that the reviewing framework and guidelines mandate such rejection.

        • So why not refuse to publish in the first place unless at least one other unrelated group has replicated? Then publish the original and the replication together?

          • Re: (Score:2)

            by imidan ( 559239 )
            That's an interesting idea, and one that would dramatically improve the reliability of the statistics published in science papers. It would also require a pretty different funding model plus a qualified group of scientists unrelated to the first group who have time to attempt replication. Ideally, replication would be done blindly, but I doubt it would be feasible that way. I think the bigger part of the problem is the money... the second team needs enough money to collect another sample and re-analyze. The

          • So why not refuse to publish in the first place unless at least one other unrelated group has replicated? Then publish the original and the replication together?

            Because both the authors and the conference organizers want novel papers. Waiting for independent confirmation of results may be in the best interests of society but is definitely against the immediate best interests of the authors and conference organizers. If independent confirmation is required, by definition novel work will be impossible to publish.

            • So run the original research and the confirmation at the same time with independent groups. Publish both results only if they agree.

          • A couple comments. In some fields (medicine, for example), it may take a long time to do a study, and replicating the study with new patients would take twice as long. It might be more useful to do a similar study with some variables changed.

            In other fields, it may be close to impossible to replicate a study. My field, for example, is linguistics; for languages that are on the verge of extinction, or in remote areas (like parts of PNG), going to the speakers and eliciting new data may be next to impossib

            • A research inaccuracy in linguistics is a problem, I acknowledge, but it is significantly less dangerous than a research inaccuracy in medicine or ecology or material science. The place where the replication crisis is greatest is in the highest stakes fields, where there's lots of money flowing around, and lots of gain (reputation and financial) for being first to publish. And they're all areas where we obviously can do replication because their whole point is to replicate the findings of some bit of resear

          • Re: (Score:2)

            by shanen ( 462549 )

            Insofar as it's a solution approach, I like it, but I agree with the other replies that largely focused on feasibility. There was also an insightful comment about the difficulty of replicating some of the complicated results.

            I'm not sure if it's a feasible solution, but I think the place to focus the additional research, including some amount of replication, is areas where the results of various studies seem inconsistent. The problem is that anything short of a full replication is introducing new variables

      • I wonder if the replication problem has anything to do with the complexity of the replication involved.

        I mean, if my paper drops a ball 100 meters and measures how high it bounces, it's not hard to replicate.

        But some of these studies involve really complex things where just doing the replication is hard, especially if involves a lot of novel test and measurement setups, software, procedures and systems.

    • The scientific publishing machine is also pretty bad in the US and China. I've been watching it up close for about 40 years. The bottom line is that you get what you reward.

      There was a recent report that about 80% of the english language scientific biomedical literature was not reproducible. I'd cite the report but there's only a 20% chance it's right.

      It also makes citing appropriate papers for publishing rather interesting, when many of the papers on the topic are so vague there is no way to tell what the paper is actually contributing, or focuses on a narrow improvement that doesn't translate to any real world application.

    • Re: (Score:2)

      by Nikkos ( 544004 )

      Gift-Authorship seems to be rampant in the US as well. A published paper that I and 2 other graduate students did has at least 4 other names on it, one of the names is the professor of the class we did the work in, the others are just friends of the professor, and one undergraduate who 'helped' by downloading the survey data.

      I got my master's degree and got the fuck out of academia. I'm making twice as much as anyone in my cohort stupid enough to stay in academia.

  • Corruption in Russia (Score:5, Funny)

    by Freischutz ( 4776131 ) on Wednesday January 08, 2020 @12:59PM (#59599462)
    Corruption in Russia,? We are all very surprised to hear this ...

  • Only Russia? (Score:1)

    by Anonymous Coward

    The West isn't immune to these very human foibles. Especially in the medical field, which is not a science.

  • Poor wording (Score:3)

    by 93 Escort Wagon ( 326346 ) on Wednesday January 08, 2020 @01:34PM (#59599592)

    "Academic journals in Russia are retracting more than 800 papers following a probe into unethical publication practices by a commission appointed by the Russian Academy of Sciences (RAS)."

    I'm guessing it's not the commission which has been guilty of the unethical publication practices...

    But then I'm old, I remember when a publication such as Science would have editors on staff.

  • Link: Report documenting the problems and journals' responses to them [kpfran.ru], in Russian

    Google Translate no longer translates web sites, but Yandex [yandex.com] does. You have to enter the web site URL by hand though.

    Unfortunately the Internet Archive doesn't want to save the Russian URL. Maybe there is a similar archive service that will?

    • I mean, they're just following the advice the great Russian mathematician Nicolai Ivanovich Lobachevsky... [youtube.com]

      But... the phrase "don't shade your eyes, plagiarize" doesn't rhyme in Russian.

      ne zakryvay glaza, plagiat

    • Re: (Score:2)

      by HiThere ( 15173 )

      Tom Leher was guilty of slander and libel for that song. Yes, it was a good song, per se, but it was factually totally biased and incorrect. And it said a lot more about him than it did about Lobachevshy.

      I liked the song as a teenager, but being catchy has nothing to do with being true.

      • Re: (Score:2)

        by Nikkos ( 544004 )

        "Tom Leher was guilty of slander and libel for that song."

        Completely false. Perhaps you'd like to try googling before you spread fake news?

    • Re: (Score:1)

      by qaz123 ( 2841887 )
      This song is factually wrong.
      The British mathematician and philosopher William Kingdon Clifford wrote in his Lectures and Essays, "What Vesalius was to Galen, what Copernicus was to Ptolemy, that was Lobachevsky to Euclid.

  • Some of these "practices" are also pervasive in the US. A lot of the time when you see a professor as the co-author on a PhD student's paper, that's de-facto "gift authorship" as well.

Slashdot Top Deals

egrep -n '^[a-z].*\(' $ | sort -t':' +2.0

Close