Wednesday, August 12, 2015

Open peer review at Collabra: Q&A with UC Press Director Alison Mudditt

Earlier this year University of California Press (UC Press) launched a new open access mega journal called Collabra. Initially focusing on three broad disciplinary areas (life and biomedical sciences, ecology and environmental science, and social and behavioural sciences), the journal will expand into other disciplines at a later date.

One of the distinctive features of Collabra is that its authors can choose to have the peer review reports signed by the reviewers and published alongside their papers, making them freely available for all to read — a process usually referred to as open peer review.

This contrasts with the traditional approach, where generally the reviewers names are not disclosed to the authors, the authors names are not disclosed to the reviewers, and the reviewers reports are not made public (commonly referred to as “double-blind” peer review).

Since Collabra is offering open peer review on a voluntary basis it remains unclear how many papers will be published in this way, but the signs are encouraging: the authors of the first paper published by Collabra opted for open peer review, as have the majority of authors whose papers are currently being processed by the publisher. Moreover, no one has yet refused to be involved because open peer review is an option, and no one has expressed a concern about it.

Collabra’s first paper—Implicit Preferences for Straight People over Lesbian Women and Gay Men Weakened from 2006 to 2013was published on 23rd July, and the reviewers’ reports can be found here.

So how does open peer review work in practice and what issues does it raise? To find out I emailed some questions to UC Press Director Alison Mudditt, whose answers are published below.
Alison Mudditt

RP: Presumably both the author and all the reviewers have to agree to open peer review before Collabra can publish the reviews? What percentage of the papers it publishes does Collabra expect will have the reviews published alongside?

AM: Authors choose open peer review as an option upon submission, so it is always their decision and as such they have already agreed in advance. Reviewers are made aware that authors have chosen this option and could opt to decline the review if they are unwilling to have their review comments made publicly available.

As a secondary option, whether or not open review has been chosen by the author, reviewers can sign their reviews. So it is possible to have reviewer comments be open, but the identity of the reviewer remain anonymous. Or, for that matter, have closed review, but reviewers sign their reviews. This is all described here.

With only one published article it is hard to project what the percentage will be, but at this point the majority of authors—for the papers currently being processed in our system—have opted for open review.

We are not targeting certain percentages, but rather want to put new options in front of people, especially given the numerous critiques of traditional closed peer review systems. This will not be for everyone, but we believe there’s much to be learned from experimentation with new models.

RP: Will Collabra make any effort to seek out reviewers who are comfortable with open peer review?

AM: The academic editors are selecting reviewers, and their top consideration will of course be the reviewer’s expertise for any given paper.

We make all the options and elements of Collabra clear when inviting external editors to be involved. Some editors are particularly interested in the open review option, and other editors have not commented on it.

No one has refused to be involved because it is an option or expressed a concern about this option.

RP: I assume that not all the correspondence is shared when Collabra publishes the reviews, and perhaps they might be edited in some way first (at least sometimes)? If so, what considerations/editorial rules are applied before making reviews public?

AM: Currently, the “open review file” is constituted by the reviewers’ comments on the reviewer form, the editor’s comments to the author based on the reviewers’ comments, and the author’s response—all as captured in our editorial system.

It is clear on the review form that there is an area for confidential comments to the editor that would not be shown to the author nor included in the openly available comments. But, for the remainder of the form, it is made clear that comments may be seen by the author and used without editing.

What is not currently shown is any earlier version of the paper and any comments or tracked changes on that. We will continue to monitor this policy and will consider other options, if it seems that useful or important elements are being omitted by not including earlier versions/changes.

And, obviously, if any discussion occurs outside of the editorial system between a reviewer and an editor, that will not be captured.

As regards editorial rules and considerations for any edits or omissions, we would discuss that with the editors as they came up. It is hard to say in advance what that might be (other than any information which is confidential and not even being revealed in the paper), so we’ll deal with that on a case by case basis.

Naturally, we would opt to be transparent about this happening should it occur beyond normal confidentiality considerations. For now we will see how it goes with it’s being clear on the form that comments may be used as written.

RP: Having started down this road (and so given concentrated thought to the matter), what would Collabra say were the pros and cons of open peer review?

AM: Speaking on behalf of UC Press (I’m not sure it’s appropriate to speak as “Collabra” in this context), we think that the inner workings of the peer review process are, purely and simply, interesting for any reader, but in particular for people who would like to see more transparency in this process.

There is clearly an argument to be made that making things open (rather than, for example, the double blind process) will help to reduce biases, problematic opinions, or hierarchical sensitivities that can affect the review process.

Equally importantly, open review starts to demonstrate the value added by the review process and to recognize the contributions of reviewers to scholarship

Finally, we all know that traditional peer review has not put a stop to whole disciplines being rocked by scandals of fabricated data and unquestioned results, and it’s possible that open peer review will actually help to improve the scholarly record.

On a related note, one of our other aims with Collabra is to get rid of the phrase “peer review lite” which has plagued the type of review that Collabra (and other OA titles) employs.

We characterize our review criterion as being “selective for credibility only”—checking for the scientific, methodological, and ethical rigor of a paper, and removing, as much as humanly possible, more subjective reviewing criteria for novelty or anticipated impact. Open reviews will support this mission—to show that there is nothing “lite” about this kind of review (and in fact, sometimes quite the opposite).

It’s too early for us to be able to identify specific problems with open peer review for Collabra, although we are aware of studies suggesting that it may be harder to get reviewers and it may lengthen the review time. Our limited experience so far does not support either of these concerns.

The other cons of open peer (as opposed to double blind) review are clearly to do with concerns about bias, the highly variable nature of peer review, and the additional costs it could impose on an already overtaxed system.

For example, a reviewer might be worried about openly and critically reviewing a more senior author and believe there could be a negative effect on her own career.

Our hope is that a more open system will improve the integrity of the peer review process, but the reality is that any system will be subject to the biases of human nature—we just think that this is more likely to be surfaced through greater transparency.

RP: Does Collabra think that there are occasions when open peer review is inappropriate? If so, when and why?

AM: Anything raised in peer review of a confidential nature which does not make it into the published article should be carefully removed from any open peer review comments that get published during open review.

That said, we (UC Press) are not really the drivers of how open peer review will evolve in Collabra or elsewhere. Since Collabra works only with external editors, editorial policies should emerge that are firmly based on the standards of each research community that publishes in Collabra.

If a community-driven majority standard emerged which stated that, in certain situations, open peer review was inappropriate, then we would respect such a decision.

RP: Are there any other learning points that have emerged as Collabra has sought to implement open peer review?

AM: It’s too early in the launch of Collabra to really be able to comment, although we have been pleasantly surprised at authors’ and reviewers’ willingness to consider the option of open peer review. That seems to be a great start for this concept.

An earlier Q&A with Alison Mudditt can be read here.


Stevan Harnad said...

So many natural questions come to mind; here are a few:

1. This is a new start-up journal, and a megajournal, and an OA journal: How representative of anything is the initial uptake rate for peer review: for quality? for scaleability? for sustainability?

2. On what evidence is it stated that most peer review (with established journals) involves author anonymity and referee anonymity? (My own guess would be that most is just optional referee anonymity.)

3. The purpose of referee anonymity is frank reviewing witthout risk of retribution: How does open review ensure this (rather than the opposite)? (Should all election voting be open too?)

4. The purpose of author non-anonymity is to allow referees to take into account the author’s track record in evaluating new work: What evidence is there that quality control is as good as or better than the current level when track-record is withheld?

5. Collabra is a “megajournal,” covering many fields: Is the answer to these questions likely to be the same for all fields?

6. Individual journals, too, have track-records for quality-control standards. How does one determine the track record of a megajournal: the average across all fields?

7. Why open peer review rather than open peer commentary following peer review, revision and acceptance? (The referees can be invited to comment openly too: I always did this when I edited BBS. And I've never chosen the anonymity option when reviewing for any journal.)

On the other hand, offering the referees of accepted papers the option of being named as the referees sounds like a good option, with no down side.

Neil Blair Christensen said...

Hi Stevan,

Initial uptake on Collabra is naturally not representative of thousands of journals; so simply think of it as feedback on the initial uptake of Collabra. Don’t to read more into it. We’re just sharing our experiences prompted by Richard’s interest. At UC Press we remind ourselves that even established and proven models aren’t necessarily representative of things that are desirable, sustainable, or scalable – after all, that’s why we pursue change.

We’re not trying to force a single model on our authors, editors, and reviewers. So, we have deliberately set up a diverse model that gives people the choices of traditional single blind peer review or various degrees of open peer review. Over time this model may very well change, as all things should. We try not to predict or theorize. We simply give people options to make their own decisions, and we share what we see.

Is it perfect? Unlikely. Was it perfect prior? While traditional peer review systems have served us well, we can all point to not only high-profile failures but also the ongoing pressures on the system. Evolution is imperfect and continuous, and we’re interested to see if some of the features we’re offering with Collabra can move us forward. Has Collabra hit a nerve with some? Likely. Has it hit a nerve with all? We doubt it..

We’re not searching for a universal solution (indeed, we think that would cause harm by overlooking the unique needs of different disciplines and scholarly communities), but if we were to hedge our bets, we suspect to find a net benefit to all in an open networked approach over a closed approach, imperfections included. After all, Collabra is an open access collaboration initiative.

Have a great weekend all!!

Neil Christensen, Director of Digital Business Development, UC Press