Pharmacology Biochemistry and Behavior had a curious retraction notice in September that we’re just now getting around to, but we think you’ll find it to have been worth the wait.
The article, “Interaction of Somatostatin Receptor-2 and Neuropeptide Y Receptor-1 in mice dorsal root ganglion neurons on the Pinch-Nerve injury model,” came from a group in Harbin, China, and Frieburg, Germany, and was published in April 2013.
According to the notice:
This article has been retracted at the request of the author because Figure 3 was inadvertently included in the submitted manuscript and is not suitable in this context.
That sounds like something a lawyer’s lawyer might have written. What’s puzzling, of course, is why the journal would have felt it had to retract the paper because of a merely unsuitable image — as opposed to, say, a manipulated one.
Let’s perform a thought experiment: Assume Figure 3 was legit but simply incorrect (either because it was from another experiment or because it had an honest error). Then pretty clearly correction is warranted, and context is irrelevant. So it was, for example, with this article from the journal:
The publisher regrets that there was an error in Fig. 2 of the above paper when originally published. The corrected figure is reproduced below.
Or this article:
The authors regret that some of the data names in the Figs. 1 and 2 are incorrect.
The correct version of the figures is as follows.
Or this one:
The author regrets to inform that a faulty figure 7 was submitted in the article on page number 466. The correct figure is provided below.
Or, even this one:
The authors regret that the above article contained mistakes in the figure legends of Figs. 2, 3, 4 and 5 when originally submitted and published. The corrected captions are reproduced below:
Fig. 2. Mitochondrial respiratory chain complexes I, II, III and IV activity in prefrontal cortex, hippocampus, striatum and amygdala in reversal model. (n = 6 for each group). Data were analyzed by two-way analysis of variances followed by Tukey test when p was significant. Values are expressed as mean ± S.E.M. *p < 0.05 difference of Sal + Sal group. #p < 0.05 difference of d-AMPH + Sal group. Bars represent means; error bars represent standard error of means.
Fig. 3. Creatine kinase (CK) activity in the amygdala, prefrontal cortex, hippocampus and striatum of rats following reversal treatment (n = 5 for each group). Data were analyzed by two-way analysis of variances followed by Tukey test when p was significant. Values are expressed as mean ± S.E.M. *p < 0.05 difference of Sal + Sal group. #p < 0.05 difference of d-AMPH + Sal group. Bars represent means; error bars represent standard error of means.
Fig. 4. Numbers of crossings and rearings in prevention model (n = 12 for each group). Data were analyzed by two-way analysis of variances followed by Tukey test when p was significant. Values are expressed as mean ± S.E.M. *p < 0.05 difference of Sal + Sal group. #p < 0.05 difference of d-AMPH + Sal group. Bars represent means; error bars represent standard error of means.
Fig. 5. Mitochondrial respiratory chain complexes I, II, III and IV activity in prefrontal cortex, hippocampus, striatum and amygdala in prevention model (n = 6 for each group). Data were analyzed by two-way analysis of variances followed by Tukey test when p was significant. Values are expressed as mean ± S.E.M. *p < 0.05 difference of Sal + Sal group. #p < 0.05 difference of d-AMPH + Sal group. Bars represent means; error bars represent standard error of means.
On the other hand, if the image was bogus, then it would be unsuitable — but unsuitable in any context, excepting perhaps a gallery of fabricated data.
We’ve tried to contact the editor of the journal for an explanation, and will update with anything we learn.
With the new year coming, perhaps editors could make a resolution to eschew such mealy-mouthed retraction notices and print only meaningful statements. But we’re not going to hold our collective breath.
“High magnification pictures show a more clearly colocalization of SST2 and Y1.”
(from Figure legend 3).
In looking at these pictures, they claim to show colocalization in the figure legend but appear to show the exact opposite in reality. If these figures were in fact A) not manipulated and B) actually the correct images of what they claimed to be, then C) it would not be suitable to show describe them as showing colocalization in this context.
Linked to figure manipulation: I wish to congratulate Springer on a retraction that has recently appeared for a paper that had an erratum until recently, not because it is nice to have literature retracted, because Springer has done the right thing to correct the plant science literature:
Erratum: http://link.springer.com/article/10.1007/s10725-013-9831-6. The notice stated “Due to an institutional conflict on research program, the corresponding author of the article would like to remove
the fig 1 from the original publication of the article.” I had queried Springer and the then Editor-in-Chief, Prof. Johannes Van Staden, how a figue could be retracted from an article without the actual article itself being retracted. After about half a year, it appears as if Springer reached the same conclusion that I reached, that this duplicated figure should result in a retraction of the paper.
Recent retraction of the paper: http://link.springer.com/article/10.1007/s10725-013-9801-z
“This article has been retracted at the request of the Publisher due to a violation of Springer’s publishing integrity. The figure 1 of the article has been duplicated from different research papers and led to some serious scientific flaw in the article.”
Springer should receive kudos for this retraction.
What is important now is for papers that referenced this fraud (figure duplication) to now request the journal or publisher to issue an erraturm to correct the reference lists. Post-publication peer review is the only way to correct the errors and fraud that is plaguing the plant science literature:
http://www.frontiersin.org/Journal/10.3389/fpls.2013.00485/full
However, this is not a time for celebration atr Springer. There are many papers in Springer plant science journals that are FULL of scientific gaffes, duplication, data manipulation, and the pressure is now on to see how Springer handles all of these cases. I am of the opinion, in my humblest position, that the plant science literature is in a total mess, and that scientists, editors and publishers are now fully responsible for correcting the errors of the past.
However, three related papers remain full of errors (scientific, linguistic and editorial), at least 50 each, and in some cases, still with duplicated figures, in which it is clear that the authors have purposefully (or really stupidly) swapped photos between 4 submissions, claiming that at least 3 different species all corresponded to the same photos. Incidentally, all four journals, including the now retracted paper, are Springer journals:
1) Mohanty P, Das MC, Kumaria S, Tandon P. 2012. High-efficiency cryopreservation of the medicinal orchid Dendrobium nobile Lindl. Plant Cell, Tissue and Organ Culture 109: 297–305.
http://link.springer.com/article/10.1007%2Fs11240-011-0095-4
2) Mohanty P, Das MC, Kumaria S, Tandon P. 2013a. Cryopreservation of pharmaceutically important orchid Dendrobium chrysanthum Wall. ex Lindl. using vitrification based method. Acta Physiologia Plantarum 35: 1373–1379.
http://link.springer.com/article/10.1007%2Fs11738-012-1163-z
3) Mohanty P, Nongkling P, Das MC, Kumaria S, Tandon P. 2013b. Short-term storage of alginate-encapsulated protocorm-like bodies of Dendrobium nobile Lindl.: an endangered medicinal orchid from North-east India. 3 Biotechnology 3: 235–239.
http://link.springer.com/article/10.1007/s13205-012-0090-4
No errata. No expressions of concern. No retractions. No justice?
That’s what I observed. The figure seems to show lack of colocalization. Maybe the retraction notice is an attempt at “we submitted to wrong figure and can’t find the right one”.
Or perhaps even, “we could never get the data we wished for and never will”. This highlights one root of the problem: testing a hypothesis where you only “want” (that is for your CV) one answer. I put some of this down to poor training.