“13 Photoshop Fails You Won’t Believe Really Happened!” A familiar tag-line on the internet, and you yet again waste a few minutes as you click through a series of impossible images of the human form. The word Photoshop, as for Xerox, has lost its Trademark status through the compliment of being incorporated into everyday English: we all know what “photoshopped” means, and it is affecting science in ways that are considerably less amusing. Now we routinely find publication of scientific figures that have been photoshopped in the most blatant way. Duplicating and removing gel bands, copying microscope images, cut-and-pasting two dimensional flow cytograms, deleting outliers: no data format appears to have escaped this wave of manipulation, even extending to complete fabrication from scratch. Why is this?
Publish or Perish
Part of the problem is related to the current way in which journal articles are selected for publication. Most journals employ pre-publication anonymous peer review, using editorial boards to implement the review process. These editors identify primary reviewers, and shepherd the submitted article in a timely manner through review. Recommendations for changes are discussed with the authors, and a publication decision is made. It is generally accepted that peer-review works best if it is anonymous. Despite efforts, there are still problems associated with pre-publication peer review, and unscrupulous scientists can take advantage of these problems. Major problems include:
- Inappropriate selection of reviewers. Authors, depending on the journal, can be permitted to suggest reviewers, and alternatively request that some individuals be excluded from reviewing. Although authors are required to declare conflicts-of-interest in selecting reviewers, this declaration may not be truthful.
- Lack of reviewer expertise in general.
The temptation for authors to cut corners (lack of repetitions, controls, non-optimal experimental designs, selective use of data, and so-on) is strong, and the way that this can lead to whole-scale, routine data manipulation is sadly obvious. Opportunities have even been exploited by some authors to provide completely fabricated reviewers and reviews. For instance, there was a case where an author admitted to providing fake reviewer suggestions – that meant he submitted reviews on his own paper using pseudonyms.
The questions become (1) how can we counteract an increasing level of fraud within scientific publications, and (2) what is the best and most efficient way to identify these articles and remove them, and their influence, from the published literature?
How You Can Help
1. Participate in Peer-review
Peer review doesn’t typically provide monetary compensation. But it is an important part of the academic workload since it aims to ensure that manuscript get a high-level of constructive criticism – the peer review system relies on obtaining high-quality feedback from experts in the field in order to allow editors to make informed and correct decisions on whether or not to publish the submitted manuscript.
2. Ensure Your Submitted Work is of a High Standard
Firstly, it is the responsibility of all authors to be certain that the work they are putting their name to is sound. All authors should see raw data, and if there is suspect data, this should be handled before the article gets submitted. Be aware that post-publication peer review will almost inevitably detect fraudulent activity. Secondly, some journals require responsibility statements indicating the roles played by all authors. If these are not specifically required, it is still good practice to include such a statement under Acknowledgments.
3. Participate in Post-publication Peer Review
Post-publication review is nothing new. We have all attended scientific seminars, journal clubs, conferences, and workshops in which discussion of published work is center stage. As journals have moved to an on-line electronic format, the opportunity for comment is starting to appear (see for example www.plos.org). Example of post-publication review sites include:
- Publicly funded sites that index journal articles such as PubMed.
- Private indexing services (requiring subscriptions) providing post-publication reviews such as Faculty of 1000.
- PubPeer. This site provides a particularly vigorous anonymous forum for comments directed at specific publications. It notifies authors of comments to allow their responses, really opening up dialogue on published papers.
- Retraction Watch takes a more journalistic approach, providing articles that describe largely the endgame (i.e. retraction) accompanying exposure of fraud or other forms of malfeasance. Individual comments are allowed on subjects of these articles.
There is also of course the avenue to go directly to the journal. In cases where readers who are experts in their field have serious concerns regarding a paper’s methods, results or conclusions, many journals offer the avenue of comments (for example, in Science, comments can be submitted concerning any article). Directly highlighting serious issues in papers serves to inform the journal, and also allows the authors to respond to the points raised. In the case of Science, anonymous comments are not permitted, which raises the issue of potential retribution in response to criticism, a particularly significant concern for early career scientists.
So What Should You do to Get Involved with Post-publication Peer Review?
The best way to appreciate these sites, of course, is simply to read them. At some point, you may become comfortable to start providing comments (all the usual rules of internet etiquette apply, no trolling, flaming, sock-puppetry, and so-on). If you are an expert, you should obviously use your expertise to speak up about manuscripts, and not just from the negative point of view.
Why You Should Get Involved
Aside from ensuring the high standard of published papers, post-publication peer review provides a great avenue for improving your own skills to critically analyze papers – initially just reading those post-publication reviews posted by others will help you to get a better sense of what to look for when reviewing papers and improve your ability to think critically about published work.