Peer review is at the heart of the scientific process. As I have written about before, scientific results are deemed publishable by top journals and conferences only once they are given a stamp of approval by a panel of expert reviewers (“peers”). These reviewers act as a critical quality control, rejecting bogus or uninteresting results.
But peer review involves human judgment and as such it is subject to bias. One source of bias is a scientific paper’s authorship: reviewers may judge the work of unknown or minority authors more negatively, or judge the work of famous authors more positively, independent of the merits of the work itself.
Double-blind: Authors are blind to their reviewers, who are blind to authors
The double-blind review process aims to mitigate authorship bias by withholding the identity of authors from reviewers. Unfortunately, simply removing author names from the paper (along with other straightforward prescriptions) may not be enough to prevent the reviewers from guessing who the authors are. If reviewers often guess and are correct, the benefits of blinding may not be worth the costs.
While I am a believer in double-blind reviewing, I have often wondered about its efficacy. So as part of the review process of CSF’16, I carried out an experiment:[ref]The structure of this experiment was inspired by the process Emery Berger put in place for PLDI’16, following a suggestion by Kathryn McKinley.[/ref] I asked reviewers to indicate, after reviewing a paper, whether they had a good guess about the authors of the paper, and if so to name the author(s). This post presents the results. In sum, reviewers often (2/3 of the time) had no good guess about authorship, but when they did, they were often correct (4/5 of the time). I think these results support using a double-blind process, as I discuss at the end.
Filed under Process, Science
As I have written previously, academic computer science differs from other scientific disciplines in its heavy use of peer-reviewed conference publications.
Since other disciplines’ conferences typically do not employ peer review, results published at highly selective computer science conferences may not be given the credit they deserve, i.e., the same credit they would receive if published in a similarly selective journal.
The main remedy has simply been to explain the situation to the possibly confused party, be it a dean or provost or a colleague from another department. But this remedy is sometimes ineffective: At some institutions, scientists risk a poor evaluation if they publish too few journal articles, but they risk muting the influence of their work in their own community if they publish too few articles at top conferences.
The ACM publications board has recently put forth a proposal that takes this problem head on by formally recognizing conference publications as equal in quality to journal publications. How? By collecting them in a special journal series called the Proceedings of the ACM (PACM).
In this post I briefly summarize the motivation and substance of the ACM proposal and provide some thoughts about it. In the end, I support it, but with some caveats. You have the opportunity to voice your own opinion via survey. You can also read other opinions for (by Kathryn McKinley) and against (by David S. Rosenblum) the proposal (if you can get past the ACM paywall, but that’s a topic for another day…).
Update: PACM has been approved, as has a new journal series called PACM PL that will collect papers accepted by major SIGPLAN Conferences. It will debut during late 2017.
Filed under Process, Science