Peer reviewing is a quasi-standard in research publishing; a synonym for publishing good, true data, based on a sound analysis, drawing appropriate conclusions, to extend pretty well into the bunch of already published research.
O.k. that’s naive. There’s a bunch of well formulated critique on the forces that work on and bias the review processes, for example the ones in the case of Top-Tier Journals: Aarssen et al. show up in their worth reading PlosOne-paper from March 2009 how publishing in such journals affects reviewer behaviour.
On genomesunzipped.org Joe Pickrell makes a strong statement, that “it’s simply absurd that our way of achieving this goal (I.e. filtering papers based on quality, interest to a community, and reproducibility) involves a handful of unaccountable, often anonymous, reviewers and editors, and takes so much time and money.”
After reasoning his critique and defining the mechanism which bias the research publishing he proposes some features for a different publishing process:
1. “Immediate publication without peer review.”
2. “One-click recommendation of papers” (like the feed in reddit or Facebook, Google+).
3. “Connection to a social network. …do people I trust like the paper?”
4. “Effective search based on the collective opinion on a paper.” Here Pickrell focuses on the community, not on the typical measures like (journal) citations. New papers have zero citations, because citations take time! Much faster and more appropriate for new literature is sorting based on the criteria in steps 2 and 3.
Yes, I like it… read it:
Why publish science in peer-reviewed journals?