What can be said that hasn’t already been said about preprints? If you’re new to this topic, I’ll wait while you browse around at ASAPbio and Biorxiv.
To move forward from discussions about the pros and cons of preprint adoption for rapid scientific communication (as it exists today), I wanted to shift the discussion to some ideas I’ve had about how we can start to make a difference in preprint adoption or address some of the concerns. Apologies if you’ve independently arrived at these or I’m completely ignorant of other efforts along these lines.
Problem 1: Not enough feedback on posted preprints.
One clear benefit of preprint submission is that manuscripts become fully and freely available when authors are ready for them to be seen. Ideally, this would enable broader, more open review by the scientific community and generate feedback that could improve the manuscript. This would theoretically provide benefits compared to a system in which only 2-3 reviewers anonymously review a journal submission. The issue is that many preprints have very few, if any, posted comments. This existing state of preprint commenting/review then does not actually come with the benefit of significant feedback and highlights that preprints have not been subjected to review of any kind.
Potential Solution: Integration of preprints in conference proceedings.
One potential solution to this is the acceptance of preprints by conferences and meetings along side poster and presentation abstracts. If preprint abstracts are made available at conferences with links to full preprints, this could draw a much bigger crowd to those that are qualified and interested in reviewing/commenting on them. This wouldn’t necessarily require much additional infrastructure or work on the part of conference organizers as abstracts are already made available. A link to preprints within abstracts would be the simple version. To go a step further, meetings could include additional sessions for preprint discussion forums or include a preprint comment-a-thon where interested parties could directly submit feedback.
Problem 2: Many people use journal branding and journal impact factor as a quality proxy and believe lack of such branding will make it difficult to know which findings are important.
Some believe that if a paper is published in a higher impact factor journal, it is more likely to be important or sound. I won’t go into the many problems with this, but for these purposes, I will instead accept this as something that is widely believed. Even if not an intentional or conscious belief, I will operate under the assumption that many have implicit biases in favor of results published in higher impact journals.
Potential Solution: A hybrid model allowing submission of preprints along with comments and reviews of preprints to journals.
For submission to journals on a “preprint track,” authors would
- post preprints,
- personally request reviews from people they likely would have requested as reviewers upon journal submission anyway,
- modify the manuscript based on reviewer comments and
- submit the final product to the journal along with reviews to demonstrate concerns from respected scientists in the field have been addressed.
The journal editor would make a determination to publish solely on the basis of appropriateness for journal and lack of conflict of interest by submitted reviewers. As journals would not require as many resources for this type of review, “preprint track” publication costs might be reduced compared to traditional review. One very attractive aspect of this is that journals with a preprint track could now compete for reviewed preprints, putting power back in the hands of scientists. So why would publishers voluntarily surrender this power? I have already heard of examples of journals approaching preprint authors to request submission to their journal. Journals may want to pick up manuscripts that might not otherwise be submitted to their journal and publish it at lower resources/cost. A preprint track at journals could bridge the chasm between preprints and journals and allow for streamlined review that still has the journal stamp of approval many seek.
Problem 3: Slow/weak preprint adoption by reluctant or unaware scientists.
Many people still don’t know that preprints exist or are unsure about whether they should use them.
Potential Solution: Get trainees involved.
Technology is one area in which adoption spreads like wildfire starting with younger users. Many of us are involved in training undergraduates, graduate students and postdocs. Introduction of preprints into the normal diet of trainees may drastically change how they are viewed in the future since today’s students are tomorrow’s scientists. Also since there are so many more trainees than there are PIs, word can spread much more quickly to PIs through their trainees.
One skill that is almost universally taught is how to critically read a scientific paper, and this skill can be taught exclusively using preprints rather than journal articles. The graduate students in my home department are required to take an “Analysis of Scientific Papers” course. The instructors of this course rotate through our faculty and each instructor choses their own theme or framework. It just so happens it is my turn to run this course in the upcoming semester. I plan to assign only preprints for these weekly courses and make a collective comment/review of the selected preprints a requirement for the class. I have recently heard of others that have preprint journal clubs and this seems like a great way to ensure that students are learning how to critically read without a safety net of a journal rubber stamp. It also allows them to become more familiar with preprints and more comfortable contributing to public scientific discussion.
These are just some of the things I’ve been thinking about and I look forward to hearing more ideas as the discussion continues.