GPeerReview – Is Google Changing Science?

Before launching into what GPeerReview actually is, let’s all rest assured that although Google has changed the world, it is not yet changing the way science is reviewed and recorded.

What is Peer Review?

In the scientific community, a similar idea to ‘page rank’ has existed for many many years. When a research project pays off and produces some interesting results, the results are often published in scientific journals such as Nature or Science. There are thousands of smaller journals out there, and each one has it’s ‘ranking’, so to speak. When getting in a large influential journal such as Nature, it’s like hitting the number one spot in a Google search.

What does all this have to do with GPeerReview? Well, in science nothing gets published without being reviewed by other experts in the field. I.e. there is a peer review process. Now, GPeerReview promises to do for scientific publishing what the world wide web has done for media publishing.

What does GPeerReview really entail then?

GPeerReview is, in short, a command line tool that allows people to post reviews of some one’s work. These reviews then get linked to the work and are publicly available.

It is not actually a tool developed by Google, but it was developed using Google Code.

GPeerReview is not just intended as a review system, but also a way of publishing new scientific work without having to worry about a journal accepting or rejecting your paper, or worrying about page numbers and other editorial limitations.

Some concerns about GPeerReview

I can’t help but have some reservations about this program, although the concept seems to be entirely well intended. As a researcher myself, I really worry about:

• Authors still have to approve  review, so although bogus reviews can be avoided, it also allows an author to dismiss a good and valid review that is critical of the paper.
• People could artificially increase their number of reviews, and there is no real way this seems to be controlled for.
• Papers that are poorly written, and would never be accepted by a conventional journal, can now ‘clutter’ the service.
• Papers go through a peer review process, but it does not seem to entail a editorial process. Editors do exist for a reason, and they often help improve a paper before it is published. This again goes back – albeit in lesser degree – to my concern of bad papers flooding the community.

• Anyone can write a review, even people who have no knowledge of your field of study. When a scientist dedicates decades of his life to a particular sub field of a sub field, then the last thing he may want is to read through potentially hundreds of reviews that are written by people who have limited or no knowledge of the field. Ultimately, reviewing the reviews is also time consuming! It’s hard enough doing it the conventional way, when a publisher only selects a handful of peers to review a paper.

Do you have any thoughts on this? Whether you’re a scientist or not, do you think this will end up being a good use of Google Code? It has caused some debate already in the science community.