Tag: STATISTICS

Why do researchers mistakenly publish in predatory journals? How not to identify predatory journals and how (maybe) to identify possibly predatory journals. Fake editor, Rehabbed retraction, Peer reviewer plagiarizing. Writing for a lay audience; Proof to a famous problem almost lost to publishing obscurity

Margaret Winker, MDLeave a comment

PREDATORY/PSEUDO-JOURNALS

  • Why do researchers mistakenly publish in predatory journals? How not to identify predatory journals

“An early-career researcher isn’t necessarily going to have the basic background knowledge to say ‘this journal looks a bit dodgy’ when they have never been taught what publishing best practice actually looks like…We also have to consider the language barrier. It is only fair, since we demand that the rest of the scientific world communicates in academic English. As a lucky native speaker, it takes me a few seconds to spot nonsense and filler text in a journal’s aims and scope, or a conference ‘about’ page, or a spammy ‘call for papers’ email. It also helps that I have experience of the formal conventions and style that are used for these types of communication. Imagine what it is like for a researcher with English as a basic second language, who is looking for a journal in which to publish their first research paper? They probably will not spot grammatical errors (the most obvious ‘red flag’) on a journal website, let alone the more subtle nuances of journal-speak.”

How should you not identify a predatory journal? “I know one good-quality journal which was one of the first in its country to get the ‘Green Tick’ on DOAJ. I’ve met the editor who is a keen open access and CC-BY advocate. However, the first iteration of the journal’s website and new journal cover was a real shock. It had all the things we might expect on a predatory journal website: 1990s-style flashy graphics, too many poorly-resized pictures, and the homepage (and journal cover) plastered with logos of every conceivable indexing service they had an association with…I knew this was a good journal, but the website was simply not credible, so we strongly advised them to clean up the site to avoid the journal being mistaken for predatory…This felt wrong (and somewhat neo-colonial). ‘Professional’ website design as we know it is expensive, and what is wrong with creating a website that appeals to your target audience, in the style they are familiar with? In the country that this journal is from, a splash of colour and flashing lights are used often in daily life, especially when marketing a product. I think we need to bear in mind that users from the Global South can sometimes have quite different experiences and expectations of ‘credibility’ on the internet, both as creators and users of content and, of course, as consumers looking for a service.”

Andy Nobes, INASP.  Critical thinking in a post-Beall vacuum (Research Information)

  • Characteristics of possibly predatory journals (from Beall’s list) vs legitimate open access journals

Research finds 13 characteristics associated with possibly predatory journals (defined as those on Beall’s list, which included some non-predatory journals). See Table 10 — misspellings, distorted or potentially unauthorized images, editors or editorial board members whose affiliation with the journal was unverified, and use of the Index Copernicus Value for impact factor were much more common among potentially predatory journals. These findings may be somewhat circular since the characteristics evaluated overlap with Beall’s criteria and some of those criteria (e.g., distorted images) were identified in the previous article as falsely identifying predatory journals, for reasons of convention rather than quality. However, the results may be useful for editors who are concerned their journal might be misidentified as predatory.

Shamseer L, Moher D, Maduekwe O, et al. Potential predatory and legitimate biomedical journals: can you tell the difference? A cross-sectional comparison  BMC Medicine 2017;15:28. DOI: 10.1186/s12916-017-0785-9

  • From the Department of Stings: A fake academic is accepted onto editorial boards and in a few cases, as editor

“We conceived a sting operation and submitted a fake application [Anna O. Szust] for an editor position to 360 journals, a mix of legitimate titles and suspected predators. Forty-eight titles accepted. Many revealed themselves to be even more mercenary than we had expected….We coded journals as ‘Accepted’ only if a reply to our e-mail explicitly accepted Szust as editor (in some cases contingent on financial contribution) or if Szust’s name appeared as an editorial board member on the journal’s website. In many cases, we received a positive response within days of application, and often within hours. Four titles immediately appointed Szust editor-in-chief.”

Sorokowski P, Kulczycki E, Sorokowska A, Pisanski K Predatory journals recruit fake editor. Nature Comment 543, 481–483 (23 March 2017). doi:10.1038/543481a

 

RESEARCH ETHICS AND MISCONDUCT

  • A retracted study is republished in another journal without the second editor being aware of the retraction. How much history is an author obligated to provide? What is a reasonable approach?

“Strange. Very strange:” Retracted nutrition study reappears in new journal (Retraction Watch)

  • A peer reviewer plagiarized text from the manuscript under review. “We received a complaint from an author that his unpublished paper was plagiarized in an article published in the Journal... After investigation, we uncovered evidence that one of the co-authors of … acted as a reviewer on the unpublished paper during the peer review process at another journal. We ran a plagiarism report and found a high percentage of similarity between the unpublished paper and the one published in the Journal... After consulting with the corresponding author, the editors decided to retract the paper.” Publishing timing does not always reveal who has plagiarized whom.

Nightmare scenario: Text stolen from manuscript during review (Retraction Watch)

 

ACCESS

  • Instructions for writing research summaries for a lay audience. “It is particularly intended to help scientists who are used to writing about biomedical and health research for their peers to reach a wider audience, including the general public, research funders, health-care professionals, patients and other scientists unfamiliar with the research being described…Plain English avoids using jargon, technical terms, acronyms and any other text that is not easy to understand. If technical terms are needed, they should be properly explained. When writing in plain English, you should not change the meaning of what you want to say, but you may need to change the way you say it…A plain-English summary is not a ‘dumbed down’ version of your research findings. You must not treat your audience as stupid or patronise them.”

Access to Understanding (British Library)

  • A retired mathematician solved, and published, a theorum proving Gaussian correlation inequality, yet the proof remained obscure because it was published in a less well-known journal. “But Royen, not having a career to advance, chose to skip the slow and often demanding peer-review process typical of top journals. He opted instead for quick publication in the Far East Journal of Theoretical Statistics, a periodical based in Allahabad, India, that was largely unknown to experts and which, on its website, rather suspiciously listed Royen as an editor. (He had agreed to join the editorial board the year before.)…With this red flag emblazoned on it, the proof continued to be ignored.“

A Long-Sought Proof, Found and Almost Lost (Quantum Magazine)

 

STATISTICS

How are types of statistics used changing over time? “…the average number of methods used per article was 1.9 in 1978–1979, 2.7 in 1989, 4.2 in 2004–2005, and 6.1 in 2015. In particular, there were increases in the use of power analysis (i.e., calculations of power and sample size) (from 39% to 62%), epidemiologic statistics (from 35% to 50%), and adjustment and standardization (from 1% to 17%) during the past 10 years. In 2015, more than half the articles used power analysis (62%), survival methods (57%), contingency tables (53%), or epidemiologic statistics (50%).” Are more journals now in need of statistical reviewers?

Sato Y, Gosho M, Nagashima K, et al. Statistical Methods in the Journal — An Update . N Engl J Med 2017; 376:1086-1087. DOI: 10.1056/NEJMc1616211

 

 

____

Newsletter #5, circulated April 1, 2017. Sources include Retraction Watch and Open Science Initiative listserve. Providing the links does not imply WAME’s endorsement.

How does the NAS suggest journals should foster research integrity? How should one critically evaluate a manuscript (plus more fake peer reviews)? One year of ORCID IDs, a Dear Journal letter from a biostatistician

Margaret Winker, MDLeave a comment

RESEARCH MISCONDUCT

National Academy of Sciences on how to improve research integrity

A U.S. National Academy of Sciences panel calls for formation of an independent group to address research misconduct and related issues, including [quoted from Retraction Watch, U.S. panel sounds alarm on “detrimental” research practices, calls for new body to help tackle misconduct ] “misleading statistical analysis that falls short of falsification, awarding authorship to researchers who don’t deserve it (and vice versa), not sharing data, and poorly supervising research – as ‘detrimental’ research practices.”

“Fostering Integrity in Research“, from the National Academy of Sciences (free PDF download available):

The document includes 11 major recommendations; those most relevant to journal editors are pasted below (emphasis added):

“RECOMMENDATION ONE: To better align the realities of research with its values and ideals, all stakeholders in the research enterprise-researchers, research institutions, research sponsors, journals, and societies-should significantly improve and update their practices and policies to respond to the threats to research integrity identified in this report.

RECOMMENDATION FIVE: Societies and journals should develop clear disciplinary authorship standards. Standards should be based on the principle that those who have made a significant intellectual contribution are authors. Significant intellectual contributions can be made in the design or conceptualization of a study, the conduct of research, the analysis or interpretation of data, or the drafting or revising of a manuscript for intellectual content. Those who engage in these activities should be designated as authors of the reported work, and all authors should approve the final manuscript. In addition to specifying all authors, standards should (1) provide for the identification of one or more authors who assume responsibility for the entire work, (2) require disclosure of all author roles and contributions, and (3) specify that gift or honorary authorship, coercive authorship, ghost authorship, and omitting authors who have met the articulated standards are always unacceptable. Societies and journals should work expeditiously to develop such standards in disciplines that do not already have them.

RECOMMENDATION SIX: Through their policies and through the development of supporting infrastructure, research sponsors and science, engineering, technology, and medical journal and book publishers should ensure that information sufficient for a person knowledgeable about the field and its techniques to reproduce reported results is made available at the time of publication or as soon as possible after publication.

RECOMMENDATION EIGHT: To avoid unproductive duplication of research and to permit effective judgments on the statistical significance of findings, researchers should routinely disclose all statistical tests carried out, including negative findings. Research sponsors, research institutions, and journals should support and encourage this level of transparency.”

 

PEER REVIEW

  • How to critically evaluate a manuscript

At How to critically evaluate a manuscript: 12 questions you should always ask yourself (Publons), a useful general approach to peer review, but it’s missing some important points (I’m sure you can find more–add your comments below):

-Can the study design answer the hypothesis posed? (e.g., is the hypothesis a question of causality but the study design is observational?)

-Do the conclusions follow from the results or do they exaggerate the importance and implications of the research?

-What are the funding source(s) and potential conflicts of interest of the authors?

  • Fall out from fake peer reviews continues with more than 100 retractions

A new record: Major publisher retracting more than 100 studies from cancer journal over fake peer review (Retraction Watch)

 

JOURNAL STANDARDS

  • Results after one year of journals requiring ORCID IDs 

“Our 2015 community survey indicated that most researchers supported the idea of their organizations requiring the use of ORCID — 72% agreed or strongly agreed that these would benefit the global research community, 21% were neutral, and only 7% disagreed or strongly disagreed. Three quarters said specifically that it would be useful if their publisher mandated ORCID iDs.”

It Takes a Village: One Year of Journals Requiring ORCID IDs (Scholarly Kitchen)

  • Technical Image Editor wanted?

Journal of Biological Chemistry is hiring editors to manually screen images for potential manipulation or duplication, before publication.

Job alert: Biology society hiring editors to screen images (Retraction Watch)

 

STATISTICS

“Dear Journal”, from a concerned biostatistician

“The safe-conducts given by the editorial system to articles that do not disclose exact sample sizes are shocking. Science must be based on the possibility to repeat comparable designs, which obviously encompasses the use of similar numbers of observations. Sample sizes given as intervals (e.g. “n=3- 18”), inequalities (e.g. “n>3”) or absurdly nebulous sentences (e.g. “n=4, data representative of 3 rats from 2 independent experiments”) are evident obstructions to reproducibility.

Similarly, it is perplexing to notice the proportion of publications that do not clearly reveal the statistical tests used. A clear attribution of tests must be given, including the post-hoc tests used after analysis of variance. It should not be sufficient to list all statistical procedures in the method section with no indication of which test was used in which figure or table.”

Dear journals: Clean up your act. Regards, Concerned Biostatistician (Retraction Watch)

_____

 

Newsletter #7, originally circulated on April 24, 2017. Sources include Retraction Watch, Health Information for All listserve, and Open Science Initiative listserve. Providing the links does not imply WAME’s endorsement.