The first author of a highly controversial — and now retracted — paper linking body weight to integrity calls the journal’s decision to pull the article “a bitter surprise” and its handling of the article after publication “deeply unfair.”
The article, “Dishonesty is more affected by BMI status than by short-term changes in glucose,” was published in Scientific Reports in July and retracted this week. Eugenia Polizzi di Sorrentino, of the Institute of Cognitive Science and Technologies at the National Research Center, in Rome, who along with her colleagues disagreed with the retraction, told us:
I cannot hide that the request for retraction came to us as a bitter surprise. Our manuscript was accepted after two rounds of revision addressing all the comments of expert reviewers. Editors came up with such a decision following an uncommon request for a post-publication revision round.
We were extremely surprised by noticing that none of the points that were then listed as reasons for retraction were considered “major limitations” in any of the two revision rounds that our work have been diligently through during the standard revision process.
We do not necessarily agree with the critiques we have received in the post-publication reevaluation, and we would have certainly been able to respond to most of them if given a chance, had we received them before publication. In fact, receiving such comments during the initial process could have helped us improve our paper. But sadly, we were not given such an opportunity.
‘Deeply unfair treatment’
Despite being an interesting approach (journals could publish post-publication public discussion about papers), we feel that receiving comments after publication and use this for justifying retraction is at best, odd. At worst, it reflects 1) a poor review process, and 2) a deeply unfair treatment toward our work (unless the editors are keen in applying the same post-publication revision approach to all of the previously published papers in Scientific Reports).
We have carried out our work following rigorous scientific methods, presented results and supporting statistics in a replicable fashion. We thought (perhaps too naively) we could rely on the peer-review process of a top-ranking, high quality journal to assess our work. We believe we have followed every step for producing “good science”, we have been responsive to the requests of the reviewers, and the paper was accepted by the editor. We therefore believe that the responsibility of any further action should be taken by the editor, not by authors. This is the reason why we decided to disagree with editors’ choices.
Unfortunately, what last is the retraction. But I think it is somehow important to clarify the process that led to such action.
Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at team@retractionwatch.com.
Such rubbish science should have been rejected without peer review. The intention to conduct this type of “research” is insidious to say the least. There is no biological plausibility for the study hypotheses and the results of such “research” only serve to further stigmatize individuals dealing with body weight issues. And now the gall to criticize the peer review process instead of responding to the series of charges that have been leveled at the validity of this “research” if you can call it that.
“There is no biological plausibility for the study hypotheses and the results of such “research” only serve to further stigmatize individuals dealing with body weight issues.”
You appear to have a preconceived idea that the hypothesis must be wrong, which is not part of the scientific method. I have not read that paper but if it was retracted not because of fraud or mistakes than that seems to me this is retraction may be due to the politically correct fat acceptance pubmed mob. Science is science, no matter how politically incorrect the conclusions may be, and that is a bad reason to retract.
Why do you think it is garbage science? It makes perfect sense that people who do not control impulses when it comes to eating would turn out to be STATISTICALLY more likely to also not control impulses when it comes to lying.
Wow, if it makes perfect sense, it must be true!
The science is garbage not because the idea is plausible or not plausible, or because it is true or not true. The topic of the paper is irrelevant. The science is garbage because the authors’ STATISTICALLY based claims do not accurately characterize their data: the statistical tests did not actually support their claims, because they were either inappropriate or applied incorrectly. If they re-analyze their data appropriately and those results do support their claim, they are welcome to resubmit their improved work.
The journal’s name is not “Nature Scientific Reports” as named in the 2nd paragraph … But I presume that’s just want people want to call it on their CVs to get more imaginary academic points via association to its publisher, Nature Research (formerly known as Nature Publishing Group), a division of Springer Nature.
The journal name is “Scientific Reports” (ISO 4 Sci. Rep., ISSN 2045-2322).
Fixed, thanks.
Yes, i wanted to comment on the journal title used. I receive applications where candidates always mention “Nature Scientific Reports”. Also, when they publicise the paper on LinkedIn and other portals, they mention NATURE…this should be avoided..
I agree that the events precipitated reflect 1) a poor review process and 2) editors are clearly not doing their jobs.
I have reviewed for Scientific Reports before and my experience was that the Editor sat on my submitted review for 2 months, before making a first decision on the manuscript and sending the exact same comments (with no new input of his or her own) to the authors.
This is sad.
That may be true. However as a journal editor I know that at least in our automated system that sends the decision email to reviewers, to let them know of the decision, it only auto populates reviewer comments, not the editors.
I have had that kind of answer from civil servants. It is the writer of a letter who carries the responibility for its content and not the tool he uses to help him write it. Programs only do, what they are specifically told to. The omission is not a software limitation but somebody’s conscious decision.
Yes, I work with those “automated systems”, too. Depending on the system, however, editors may be able to edit (hence the term “editor”) their feedback to the reviewers. I habitually delete the version that sends the reviewer responses only and always include the full decision letter in which I try to integrate, and arbitrate between, the reviews and my own concerns & suggestions. But I realize that, depending on the system, this option may not be available in some of those manuscript-management systems. If that’s the case, then the software does not serve the editorial process and was not worth the money paid for it. The journal should get a new & better one in a jiffy!
Its kind of ironic, don’t you think?
Metabolism surely is a factor in obesity, but the role of personality in obesity may have been overlooked in recent. I don’t know, I am not an expert. This research clearly shows there is a connection between undesirable personality traits (lying) and a higher BMI in the population that has been studied. Whether this is true for people outside of the population study is a matter of genetics. The notion that someone’s appearance can be an indicator of their personality has been taboo in the past decades. This is politically dangerous territory, lest it turns out some populations are less honest than others.
Exactly. I am 99% sure that the main reason this was retracted without notification and explanation by the journal is the outcry from the hurt feelings it may have caused. I wish the scientific community could spend some effort judging papers for their content and not it’s political implications one way or the other.
Perhaps the next subject these authors study is the correlation between melanin levels and intelligence, or melanin and propensity for violence. I am very confident grant funds would pour in from National private sources.
(Sarcastically suggested from ethical and moral outrage)
Hi Traci
I am surprised you use the word outrage. I would think a more appropriate emotion is shock. I just don’t see why you would be outraged by someone conducting research on something that is a tangible, quantifiable parameter. I would expect shock as more likely than outrage when it turns out some presuppositions you have had about society has been overturned by scientific evidence. Science not righteous ……. therefore science wrong! Sorry, but that is a caricature of a cave person or a 15th century clergyman!
In any case, if we ignore that personality is one of the underlying factors of obesity, I think that could result in ineffective treatment. But this is what all thin people secretly think – why don’t they just eat less? This study suggests there may be a biological reason for this. Lower threshold for action due to natural impulses. I could see this as positive in some situations – for instance, acting on an impulse to correct someone, even if their mistake does not directly affect me. I think this kind of research would be more digestible if it were to link a negative physical trait (obesity) with a positive personality trait (for instance vigilance or spontaneity).
Best wishes,
Antoni
Not my area, but I thought that the old notion of Sheldon’s body types and personality had been long debunked, https://www.theatlantic.com/health/archive/2018/11/body-stereotypes-personality-debunked-eugenics/575041/
I’m sorry, are you saying that the most recent literature linking body type and personality is from this professor Sheldon from the 1940’s and 1950’s? If so, then it seems that it may be time to update. Anyways, links between physical and behavioral phenotype will be present in computerized medical records that are being implemented in some countries including the USA.
I’m not saying old stereotypes are true, its just that it is certainly not impossible that there is some deeper connection between weight and behavior. Researchers proved this much by showing that people who engage in spontaneous body movements, such as leg shaking, burn a significant amount of calories in the process. What may be a waste of energy in a resource poor environment could help them stay thin in our society of excess. Spontaneous twitchy motions can for instance be useful in environments where there are many insect parasites which you have to defend against. We don’t know how these things work at the molecular level yet, but it has to do with brain chemistry. These things are complicated and actually they are not at all intuitive. So, many common stereotypes may certainly not be correct.
Hi Antoni, I was merely pointing out the outcome of the earlier, very popular (at the time) theory of body types. But, in the service of full disclosure, I admit to being an ‘endomorph’ who actually believes to be endowed with some of the stereotypical personality characteristics associated with that body type, including cheating when dieting. 🙂
BTW, I enjoy your thoughtful posts.
“At worst, it reflects 1) a poor review process, and 2) a deeply unfair treatment toward our work (unless the editors are keen in applying the same post-publication revision approach to all of the previously published papers in Scientific Reports).”
1) Yes, the peer review was shoddy, which is not the authors’ fault. But that is not a reason to retract the paper. One could argue this is true of most (if not all) retracted papers. But correcting mistakes is what science is about. The field is more important than your publication record or feelings.
2) This is the equivalent of complaining that you got a ticket for getting caught speeding while other people who were speeding did not get caught and did not get tickets. Is this unfair? Yes, in the sense that in a perfect world, all speeders would be caught and ticketed. Is this unfair to you? No, because you were still speeding. What happens to everyone else is of no concern in this case. Your paper was high profile, which was clearly your intention given the provocative title, and thus more likely to attract scrutiny than most papers. Your paper was like a red Lamborghini racing down the highway. Extraordinary claims require extraordinary evidence. Your claim overreached your evidence. Do better science.
My second sentence in the first point is in error; it should be “But that is not a reason to NOT retract the paper.”
Best comment by far!
I think that at the core of the issue here is – can we call all things exactly as they are? “What is in a name? That which we call a rose by any other name would smell as sweet.” as William Shakespeare wrote in Romeo and Juliet. And yet, both of the young lovers end up dead.
I remember reading somewhere that as we get older we start to accept the deeper meaning of names – that a name in itself has some meaning. For instance that fine wine from an old winery tastes better than a sophisticated inexpensive blend. When I was younger I thought that this was stupid, that maybe this was just placebo. But I think what happens is that as we get older we make peace with certain issues, and we accept them, exactly as they were before but without the hostility and the fear. A peace is forged, and is symbolized by just one word or concept. Harmony. Justice. Truth. If the word is challenged, the underlying peace which has been struck and reinforced is also challenged.
On the other hand, science is supposed to be impartial. It is only supposed to be about hard, quantifiable data. The hard fact is that obese people in the study were more likely to cheat. Is that why the study was retracted? Well no, the reason it was retracted is that the editors came under pressure from people who have made a certain peace with obesity. Publishing that paper although it was the truth was an insult to peace. But there is no place in science for peace. These two systems are incompatible. But I believe in our society they can coexist. Facts should be sugar coated for the general audience, or filed under appropriate names to allow room for interpretation and dispute. Battles between systems of values should aim to inform and convince rather than impose. Harmony has to be reached, but truth should not suffer. In a lot of words, I have made peace with the issue. All is as it was before, in name, but there is peace to remember. A knot in a string is formed, tightened, and tightened more until it disappears. And yet it is not the same, because we remember what happened.
What is really fascinating is that all these notions about justice and truth stem from brain chemistry, and that research such as this is bringing us closer to really understanding what consciousness is and integrating that into our system of values. Some people value harmony more than truth. But without harmony truth cannot exist because without it we are reduced to animals fighting for existence, and there isn’t time to ponder deeper meaning. That is why harmony must triumph over truth in the long run, because without harmony neither can exist.
You are all over this comment section haranguing about how we as a society need to admit that fat is a moral failure (as if that’s not the status quo!), and it really makes one wonder what your stake in all this is.
From this story is appears & the comments on PubPeer it appears that the peer-review process was less than optimal in this case. That said, that does not give the Authors a ‘get-out-of-jail-free’ card. If the statistical methods really have basic flaws in the paper, then the onus is still on the Authors.
I review quite a few papers having to do with obesity. In my extensive experience, obvious flaws are often overlooked as long as the paper shows obesity in a bad light.
So what is the difference here? Is it something insidious in our politically correct society, or were the authors simply unlucky and got buried under a social media avalanche? 🙁
There is no difference. Reviewers and editors were uncritical because the paper said something bad about obese people. Thus the totally obvious flaws got completely ignored. That’s why the paper got accepted in the first place. If the paper had said that normal weight people were more dishonest than obese people, the flaws might well have been noticed and the paper might have been rejected. In this case, however, the flaws were so egregious that when pointed out, they could not continue to be overlooked.
Your own analysis here appears to be as personally biased as the bias you are accusing the editors of.
I am just commenting based on my own experience, which is pretty extensive. I have reviewed quite a few “obesity” articles with numerous flaws that I have pointed out, which lead to either considerable change in the articles or to the articles being rejected. But I am often struck that none of the other reviewers seem to have noticed anything wrong and not only just accept the results almost unquestioningly but praise the results for being important. Such articles are characteristically those that say something is ‘bad’ about obesity (like the article under discussion here, finding that obese people are more dishonest).
I haven’t read the paper but I disagree with the use of the word fair. It doesn’t matter if something is fair or not, good science can become better science.
BMI/weight and honesty sounds like an argument that could take place on a media outlet, rather than on a journal. I don’t see where you could possibly draw a causal relationship. But you may make a few observations.
The authors are obviouly aware of the “Cum hoc ergo propter hoc” rule, and clearly stated:
“Due to the correlational nature of our study, we are not able to infer causality between obesity and moral behaviour.”
Btw, one striking result of their work is, at least for me, that a lot of people are cheating, regardless of their BMI status. The paper is based on a game with rather small rewards (0, 3, 5$). I wonder what would happen with strongly unbalanced rewards, for example 0, 5, 500$, or negative rewards, like -10, 0, 5$.
Are we going to bring up the anorexics or bulimics? Statistically, if they can lie about eating, they can lie about anything. Morbidly obese people don’t lack impulse control any more than a morbidly thin person does. I think it would be worth looking into those diagnosed with impulse control (bipolar) and correlate how the need for control either increases OR LIMITS calorie intake.
Wow! This was a very spirited & stimulating comment section. I wish I was friends with all of you! I’m new to research (artificial sweeteners), but already seeing some issues with the way papers are written and quite surprised at some that slide past peer review despite significant flaws. It’s almost as if these journals are trying to compete with their less sophisticated and more provocative cavalier step siblings. I don’t think one should ever be rocked or jarred by the title of a scientific paper, unless the evidence is truly that. And even then, I’d hope that a researcher that finds a highly correlated unexpected relationship or strong evidence of novel concepts proceeds with caution and humbleness. When an author is out for personal gain, it shows in their words. You can feel the ego as you read. Thank you all for the invaluable insights!
See Aaron’s comment above.
Looks to me like these authors were not paying enough attention to the words they were using. “Lying” and “dishonesty” are loaded with negative moral meaning and are red flags in any scientific paper. All the authors would have needed to do was to reformulate their hypotheses avoiding these words and replacing them with more neutral expressions. This consideration should have already played a role in the overall design of their experiments before they even performed them. We’re certainly NOT going to correlate any kind of biological mechanism with “lying” or “dishonesty”. Lets look for a neutral (or positive) correlate before we even start.
“It’s so unfair to say no after mama said yes!!!” is absolutely the mature answer I want to hear from a young scientist who dropped a ethically questionable paper.
Yes, the review process could, and should, have been better. That doesn’t make the problems of a paper magically disappear.
Feynman talks about this in his discussion of the Challenger disaster. You can look at external testing as a way to catch your errors, so you don’t have to do it yourself. Or you can look at external testing as a final failsafe, and reason that if it catches anything you have messed up in a big way. The group that took the first approach was the one whose work failed, and he did not think this was a coincidence.
We seem to be seeing this attempted excuse — “the reviewers and editors should have caught the problems” — more and more often, and it is a very bad one. Authors, not reviewers, not editors, are responsible for catching errors. In many cases reviewers and editors *can’t* catch them, such as papers which don’t include primary data with the submission.
We also seem to be seeing an emphasis on whether the authors are at fault. It does not matter. If the paper is wrong, no matter whose fault it is, it has to be retracted: otherwise it introduces poisons into the scientific bloodstream. Retraction is not a punishment, it’s an essential cleanup.