Preprints do not promote confusion or distortion in the public understanding of science

Today, Corina Logan, Laurent Gatto and I have a short correspondence published in Nature. This is a response to an article criticising the relationship between preprints and scientific journalism. Our piece is just one of many, and we have included a list below the full response below. We welcome wider discussion on this issue!

Original piece

On 24th July, Tom Sheldon wrote in Nature News and Comment about the potential for preprints to elicit public confusion and misinformation and distort the public understanding of science. We are disappointed with the chosen narrative which perpetuates the fallacy that peer review is a guarantee for validity, whereas non-peer reviewed preprints present misinformation. There are countless examples of how peer review has failed to stop ‘bad science’ from entering the published record (e.g., Margalida and Colomer, 2016), with high-profile cases including the ‘vaccines cause autism’ (Eggertson, 2010) and ‘arsenic life’ (Hayden, 2011) scandals. Therefore, Sheldon’s article inadvertently highlights issues regarding the reliability of peer review.

Responsible and well-respected journalists rarely consult just a single source for their works; they consult multiple independent sources to verify research. This critical evaluation process is not contingent on whether research has been peer reviewed – Sheldon undermines this standard best practice for journalism and places the blame on preprints.

Sheldon states that journalists need embargo periods to have enough time to verify research findings. However, preprints, through early dissemination of research outputs, actually offer more time to gather expert feedback for journalistic vetting, and in an unrestricted manner. We expect scientifically-literate journalism to fact-check all sources, and to verify the status of the research with authors, irrespective of whether it has the peer review label. To assist this, most preprint servers either explicitly label preprints as ‘Not peer reviewed’, or have editorial ‘sanity checks’ in place to prevent the posting of junk science.

The reality is that many research articles have some errors. The identification of errors is a function of time, which preprints help by extending. We feel the need to remind readers that peer review is not a magical process that filters out ‘inaccurate’ science, and that even the peer reviewed record is imperfect. Preprints and peer review are not in competition; they are complimentary.


Eggertson L. (2010) Lancet retracts 12-year-old article linking autism to MMR vaccines. Canadian Medical Association Journal 182(4):E199-E200 https://doi.org/10.1503/cmaj.109-3179

Hayden, E. C. (2011) Open research casts doubt on arsenic life. Nature News,  https://doi.org/10.1038/news.2011.469

Margalida, A., and Colomer, M. À. (2016) Improving the peer-review process and editorial quality: key errors escaping the review and editorial process in top scientific journals. PeerJ 4:e1670 https://doi.org/10.7717/peerj.1670

Sheldon, T. (24 July, 2018) Preprints could promote confusion and distortion, Nature 559, 445 https://doi.org/10.1038/d41586-018-05789-4

Other links

2 thoughts on “Preprints do not promote confusion or distortion in the public understanding of science

Leave a Reply