When the scientific system fails: predatory journals and an example of authors not doing their job

Our current system of science publishing works something like this:

  • You design a study, collect the data and write a paper.
  • You submit the manuscript to a peer-reviewed journal.
  • An editor receives, evaluates your paper and decides if it goes out for reviewing by other people in the field, who are chosen by him/her. If he/she doesn’t think your manuscript is suitable, he/she rejects it. When this happens, one usually submits to another journal(s) until one accepts it for publication.
  • If sent for review, after some time, you receive the comments of (usually) anonymous reviewers on your manuscript.
  • If the editor choses so, based on negative reviews, your paper is rejected.
  • If there were mostly positive reviews, you are invited to submit a revised version addressing the reviewer’s comments. This basically is, sometimes, a trade-off between what the reviewers ask and what you think should be included.
  • If you make the reviewers happy, then your paper is accepted and published.

Although this is not a perfect system (read this), its how science publishing works. Peer-review serves to be a kind of “quality-control” on what is being published, although many times (also depending on the field) it fails to do an adequate job (in the end, peer-review is made ad honorem and takes time out of other duties).

I came across this study (Lopes et al., 2018), co-authored by Brad Schoenfeld, while writing for another, longer post on a study by the same subject. Now, this journal is rather obscure (never heard of it and the last article published there that is indexed on PubMed is from September 2016), but that is not the important point. The interesting part came from reading the following section of the discussion (references are linked in the cited section):

“In agreement with the current results, Rahimi (19) reported
no difference between moderate inter-set rest interval
(60 seconds, 90 seconds and 120 seconds) on the
number of repetitions completed over 4 sets of squat exercise
performed to muscular failure (85% 1RM). Furthermore,
a study conducted by Willardson and Burkett (31),
also indicated no difference between 60 seconds and 120
seconds rest intervals on the total number of repetitions
performed over 4 sets of bench press and back squat exercise
using an 8RM external load. The present study expands
on these findings to show that TLL was similar between 60
seconds and 120 seconds rest interval during a full-body resistance
exercise session (6 exercises).”

Following up on the references, Rahimin, and Willardson and Burkett, I found that the two abstracts were virtually identical:

Side-by-side comparison of both abstracts.

Willardson and Burkett was published in a well-known journal (Strength and conditioning research) in February 2005. The bottom one (Rahimi), was published in December 2005. It looks like Rahimi just copy and pasted, word for word, the entire abstract, and just deleted the mention of the bench press (and changed 8RM for 85% of 1RM). Moreover, he is the single author (unusual, to say the least), and the baseline data for participants is exactly the same as for the Willardson and Burkett paper.

But it gets worse. If one looks at the full text (here and here), everything is basically the same. He just appears to have re-written (with a lot of errors) some parts or copied exactly the same for other parts. Below a couple of screenshots, but just reading for a couple of minutes each paper is enough to notice the problems:

From the methods. Here, the baseline data is different from the copy-pasted abstract.
Apparently just changing some number is enough.
Most of the discussion is a copy-paste from the Willardson and Burkett paper.
More examples…

Judging from this, its pretty clear that the Journal of Sports Science & Medicine is a predatory journal and that the Rahimi paper is a bogus article. It is very likely that no actual peer-review took place. But for the uninformed eyes, it is a valid “scientific reference”. (Edit 14/09/18: See below).

But what is worse is that this obvious example of plagiarism went unnoticed by Lopes et al., 2018. Even in the abstract it is explicitly clear that this is a case of plagiarism and falsification. It is beyond me how the authors from this study failed to see any problem and cited them both in the same paragraph. It may be confirmation bias, as they appear to have found something similar. But overall, it is the responsibility of the authors to adequately check references and cite them appropriately. Unfortunately, reviewers don’t have enough time to fact-check all references. In this case, it was pretty obvious for me after only a couple of minutes, but in other cases it might take more effort to detect something fishy. None of the NINE authors (including Schoenfeld) in Lopes et al., 2018 noticed anything and basically agreed with what is written (apparently, I have been told that in some cases, not all authors read the paper before its submitted, which is something new for me, at least acknowledging it publicly). Moreover, not a single review was familiar with these two references or they weren’t checked. In fact, the Rahimi paper has been cited 98 times!

Bottom line

There is a lot of crap disguised as scientific research available and the problem with predatory journals (and their identification) is far from being solved. However, it speaks volumes and very negatively of authors failing to notice such an obvious case of plagiarism, despite being a very clear case beginning from the abstract. You didn’t even need to read the full text; the abstract is an identical copy. Moreover, it appears that some authors do not critically evaluate what they cite (maybe don’t even read the studies cited), or only exert selective criticism towards studies that they don’t agree with.

In conclusion, always be skeptical, critical, follow up on references and do not take anything at face value.

Edit 14/09/18

The editor-in-chief of the Journal of Sports Science & Medicine has left the comment pasted below. Perhaps I used the term “predatory” losely based on this outrageous case and the current pay-to-publish model. Based on his response (and the input from others), the J Sport Sci Med might not really be a predatory journal and this is just an unfortunate example. However, I will leave the original post as it was originally for people to read it, and then see the responses as they were developed.

On the other hand, I find the response by Rahimi completely unsatisfactory. The almost exact copy paste of the abstract (even including the baseline data from the other paper) makes it very unlikely that it was done unknowingly. Being a MSc student is also not a valid excuse for plagiarism. I find the evidence by reading the paper pretty compelling and unexcusable. You can judge for yourself.

Dear Lucas
I am the editor-in-chief of J Sport Sci Med and be awared about the matter via one of your follower.
You are right some part of the papers are identical. First, I have to inform you that it is a shame for the journal, which is indexed in top indexes and has a high IF, and I am so sorry about it.
JSSM has started publishing life in 2002 as one of the first free electronic journals on the world in the field of sports science. JSSM is a non-profit journal and there is no company and organization behind the journal. At that time, we do not have a program to check plagiarism because of the financial problem of the journal. Therefore, the manuscript was not controlled by a program for plagiarism. JSSM has started to take a publishing fee since 2010 and then the journal has a money to buy such a program of “Ithenticate” and we use it for all paper since 2010. JSSM did not earn/take any money from the authors for publishing papers or others until 2010 so we did not get the money from that paper of Mr Rahman to publish it.
When I checked the process for the manuscript in 2005 I have also had an interesting note to share with you. In 2005, JSSM was using section editor system which is still continue. The process of the mentioned manuscript has been managed by Dr William J Kraemer as a Strength/Resistance Section Editor of JSSM and accepted by him for the publiaciton (please see the board in 2005 at http://web.archive.org/web/20051208115212/http://www.jssm.org:80/bord.php). Interesting point is that Dr William J Kraemer was also the editor-in-Chief of “Journal of Strength and Conditioning Research” where Willardson and Burkett paper has been published in 2005.
Our reviewing process is very clear as see at https://www.jssm.org/newauthors.php and authors are responsible about scientific and legal responsibility of their article and sign a form about it (see Copyright Transfer Statement of JSSM[.doc] at https://www.jssm.org/newauthors.php; “We are in agreement with the statements mentioned in “Authors Instruction Form” of JSSM and the above-mentioned requirements. We accept scientific and legal responsibility of the article.”).
Each article submitted to the journal send out at least 2 independent reviewers and we use double blinded process since beginning of the journal and Rahmani’s manuscript also sent out the reviewers and accepted after revisions.
You must be sure that the JSSM has not been a predatory platforms for publishing plagiarized articles anytime. Because of the financial problems, we could not have a chance to make a plagirism control during that time. As you noted, it is also interesting that although it has been cited several times in last 13 years including review articles no one could not aware about it.
Thanks for your great work to show the plagiarism and so sorry about shameful position of our journal.
Best wishes
Hakan
p.s. I also asked to the author about the case yesterday and he has replied me immediately. Just to inform you I added the letter below.
Dear Dr Gur
I apologize for this problem and thank you for informed me. This study was conducted in 2005 and I examined the effect of three different rest intervals including 1, 2, and 5-minute on squat performance with 85% of one repetition maximum (1RM), But Willardson and Burkett (2005) examined 1, 2 and 5 minutes of rest intervals on the squat and bench press volume with an 8 repetition maximum (8RM) load. The intensity in these two studies were completely different and bench press volume examined only in Willardson and Burkett (2005).
Unfortunately, when I was starting to written my manuscript my English was poor, so I had to copy some of the text from Willardson and Burkett (2005), but I cited Willardson and Burkett (2005) in my manuscript and I thought this is enough.
It should be noted that at that time I was MSc student and I didn’t have any information about plagiarism. But now I know I made a mistake and I apologize for this problem, would you please give me a solution.
Best Regards,
Rahman Rahimi