blog

In this blog, Molly Cranston; Senior Associate Editor at F1000 and Jonathan Threlfall; Editorial Data Lead and Senior Assistant Editor, discuss the policies and guidelines implemented by F1000Research to ensure our adherence to the core practices of COPE.

As the publishing world continues to evolve, guidance on publishing standards and ethics develop and change. We believe that all publishers and journals should be held accountable to external guidelines when it comes to publication ethics, and we strive to ensure that our guidelines to authors, best practices and workflows are kept up-to-date with the latest information provided by policy leaders in the publishing field.

The Committee of Publication Ethics (COPE) is a well-established group that provides guidance and resources concerning publication ethics to all those involved in scholarly publishing. COPE provides an audit to its members detailing the ten core practices of COPE, against which guidelines and best practices can be checked. As a member of COPE, in June 2019, F1000Research decided to undertake the COPE audit to ensure that our own guidelines, best practices and workflows were compliant with the latest COPE guidance. As F1000Research is a leader in open science—we were the first publisher to launch with a mandatory open data policy and an early adopter of open peer review— we decided that in addition to the COPE audit, an audit of our policies and best practices surrounding openness and transparency should also be scrutinised against the Transparency and Openness Promotion (TOP) guidelines provided by the Center for Open Science (COS).

We were very pleased that the results of the audit showed that overall F1000Research adheres well to both the COPE core practices and TOP guidelines. The eight TOP guidelines are ranked by three levels of transparency (Level 1, Level 2, and Level 3), and the COS recommends that journals “…implement a mix of level 2 requirements for data, materials, and analytical code transparency and level 1 disclosure practices for reporting guidelines, preregistration, and replications”[1]. F1000Research scored levels 2/3 for data, materials, analytical code, reporting guidelines and replication, and level 1 for preregistration, showing that we are well versed in transparency and openness in these practices.

Regarding the COPE audit, we appreciated that there were instances where refinements and additions were required in order to abide fully to the core practices. For example, it was clear that our previous practices regarding image manipulation helped protect us from inappropriate image manipulation but were not sufficient to catch all instances of impropriety. A 2016 paper by Elisabeth Bik and colleagues[2] found that 3.8% of the large sample of papers assessed contained suspicious images, with half of these showing signs of deliberate manipulation. To try and mitigate this at F1000Research, we developed a practice for detecting potential manipulation, drawing on guidance from COPE and reading policies from other publishers, societies, councils and committees. While Bik is able to assess images by eye, we’ve decided to use Adobe Photoshop and the suite of tools developed by the US Office of Research Integrity to give us a little help with forensic image analysis. Where improper manipulation is suspected, we will follow guidelines issued by COPE. You can read about our new image manipulation detection policy in our article guidelines (section 16) and in our policies.

Since the audit, we have overhauled our policies and guidelines and implemented most of the required changes. Over the coming months, further edits on the F1000Research website will be made to ensure that we are fully compliant with the requirements set out by COPE, and we intend to carry out an audit every year to sustain our adherence. We encourage other publishers to carry out regular audits of their policies and best practices to make sure that, as publishers, we are providing correct and transparent information to authors, ensuring that trust and understanding is maintained with the researcher community.

---

[1] https://cos.io/top/. Accessed 15^th August 2019.

[2] The Prevalence of Inappropriate Image Duplication in Biomedical Research Publications. Elisabeth M. Bik, Arturo Casadevall, Ferric C. Fang. mBio Jun 2016, 7 (3) e00809-16; DOI: https://doi.org/10.1128/mBio.00809-16