Now this is transparent: Retraction for plagiarism earns 4-page editor’s note

A journal has retracted a paper about 3D imaging after concluding the authors used equations from another researcher without attribution — and has conveniently included a detailed editorial explaining exactly what happened.

It’s rare for us to see a journal be so transparent in explaining what went wrong with one of its papers, so we’re thanking Stuart Granshaw, from Denbighshire in Wales, UK, the editor of The Photogrammetric Record, for “doing the right thing.”

Even the retraction note is reasonably forthcoming:

The article ‘automated 3D feature matching’ by Tzu-Yi Chuang and Jen-Jer Jaw, published in The Photogrammetric Record, 30(149): 8-29 and placed online on 12th February 2015 in Wiley Online Library (wileyonlinelibrary.com) has been retracted by agreement between the authors, the journal Editor (Stuart Granshaw), The Remote Sensing and Photogrammetry Society, and John Wiley & Sons Ltd.

The retraction is due to the inclusion of unattributed material, especially relating to the equations, which have appeared in equivalent forms in previous publications. The basic theory behind the formulation was mainly developed and explicitly formulated in Han (2010a), Han (2010b), Han et al. (2011), Han and Jaw (2013) and Han et al. (2014). These papers already established in detail similar theory, developing comparable equations necessary for estimating the rotation, scale and translation parameters as proposed in the RSTG approach in the Chuang and Jaw (2015) paper in the The Photogrammetric Record. The use of multiple geometric features and the quality assessment based on distance and angular discrepancies have also been discussed at length in these previous publications. These papers, listed below, should therefore have been cited by the authors to attribute work already published in the literature.

References:

Han, J. Y., 2010a. Noniterative approach for solving the indirect problems of linear reference frame transformations. Journal of Surveying Engineering, 136(4): 150–156.

Han, J. Y., 2010b. A non-iterative approach for the quick alignment of multistation unregistered LiDAR point clouds. IEEE Geoscience and Remote Sensing Letters, 7(4): 727–730.

Han, J.-Y., Tseng, H.-P. and Lin, C.-T., 2011. Quality assessment for LiDAR point cloud registration using in-situ conjugate features. IGARSS 2011, July 25-29, Vancouver, Canada.

Han, J. Y. and Jaw, J. J., 2013. Solving a similarity transformation between two reference frames using hybrid geometric control features. Journal of Chinese Institute of Engineers, 36(3): 304–313.

Han, J. Y., Guo, J. and Chen, Y. J., 2014. Registration of vector maps based on multiple geometric features in topological aspects. Survey Review, 46(336): 219–230.

Let’s take a look at the four-page-long editorial by Granshaw that accompanied the paper in the same issue. In it, he explains that he received an email from a reader in June 2015, which read:

“I am a reader of The Photogrammetric Record. This publication has been a good reference for my research work. Recently, an article published by Chuang and Jaw in your March 2015 issue draws my attention. That article presents an automatic approach for matching 3D features. It is an interesting topic. However, I noticed that the presented methodology is very similar to that in previous publications (all from the same author [J. Y. Han]) without giving a proper reference. More than 60% of the equations constituting the major part of the presented methodology are duplicated from the previous publications. The basic idea and designated purpose are also the same. In my opinion, the publication of the later article is unethical and should be prohibited. It also undermines the academic reputation of this journal.”

The email went on to list four references that the reader felt should have been included by the authors.

Near the end of the same month, a second reader contacted Granshaw with similar concerns, listing many of the same missing references. Here’s some of what the second reader said:

“I always strived to combat the unethical scientific behavior of published plagiarized material. I am obliged to uncover one of such blatant cases that I just discovered in your journal.. . . the authors did not cite proper references to the equations that appeared in previous publications. Any unsuspicious reader could only interpret that these essential equations framing the proposed methodology were first and solely developed by the authors themselves. Unfortunately, this would amount to a misleading interpretation. The basic theory behind the formulation was mainly developed and explicitly formulated in [various papers published by] Han.. . . I also would like to stress that I do not consider the reviewers of your journal or any of the publications department staff responsible for not discovering at the review stage that the paper in question . . . escaped the normal scrutiny procedures.. . . In my opinion, it is the sole responsibility of the authors and only they should be blamed for embarking on this unscrupulous conduct.”

After a third reader also contacted Granshaw with similar complaints, he decided to take action. Granshaw wrote:

As the current Editor, I had never received even one complaint about a paper prior to this case, so to receive three concerning the same paper provided incontrovertible evidence that the authors needed to provide a full and convincing response to the allegations made.

In the editorial, Granshaw noted three “unusual features” of the case:

(1) The normal automated checks such as iThenticate from CrossCheck (applied to all ScholarOne submissions to the Record) check the written text but not necessarily the equations.

(2) The originally submitted paper did not include the offending equations yet did include one of the missing references. It seems that efforts to satisfy reviewer comments resulted in the final, unacceptable result. 

(3) The paper’s authors, T.-Y. Chuang and J.-J. Jaw, work in the same institution, indeed in the same department, as the non-cited author J.-Y. Han; they should therefore be fully aware of his work.

According to the editorial, the authors of the original paper have agreed to the retraction statement after “much correspondence, and with some reluctance.” 

It goes on to explain the journal’s decision further:

There is no implication that the entire paper is unworthy of publication, but that the additional references (now electronically linked to the paper) should have been cited in the version published. What the two authors have effectively stolen are the ideas behind these equations and it is hoped that linking the paper to these missing references provides at least some redress and improves the integrity of the literature. 

The editorial ends on a more positive note:

…our own bad apples are mercifully few and far between and are normally caught before publication in The Photogrammetric Record.

The 2015 paper, “Automated 3D Feature Matching,” has yet to be cited, according to Thomson Reuters Web of Science.

We’ve reached out to the authors of the study, Tzu-Yi Chuang and Jen-Jer Jaw who are based at the National Taiwan University in Taipei, and will update the post with anything else we learn.

Hat Tip: Rolf Degen

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our new daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.