A Possible Solution to the Bias Against Women in Science

What about if names are not provided at all?

Clapping

Chardyspal

reminds me of the study that showed that women musicians were 30%-50% more likely to get hired by an orchestra if they auditioned behind a screen. sounds like a good idea to me.

Wouldn't it be easier to just use initials? Like, I'd be S Coleman, at least for review. In fact, double-blind review would probably work even better.

I also remember that musicians study. I read about it in the book "Blink.' I agree that the less non-essential information--age--year of graduation--marital status--gender--ethnic affiliation...just the facts, sir, just the facts the better our selection process.

Is there any reason why a reviewer would need to know the names of the authors of the paper ? If there is not then a simpler solution might be just to withhold those from the reviewer, at least until the review is done.

i also think that reviewers should not know the name of the authors of the paper they are reviewing, but some people had objections to my proposal of double-blind reviewing (see the comments section in the linked essay in the above blog entry for more detail) -- some objections that could be thought of as reasonable. so this "gender neutral name" scenario is the solution i thought of.

the problem with using JUST initials is, unless it is the journal's policy for EVERYONE to use their initials, those authors who CHOSE to use initials were discriminated against, albeit, not as badly as those authors with obvious female first names. the scenario for favorable reviews looks like this;

stuart coleman > s coleman > susan coleman

I work in a pretty insulated community (Computer Science, "systems" research) that mostly publishes in conferences. A constant complaint is that a huge portion of the papers in our big conferences every year are by the same authors, groups, and conferences.

We've done double-blind reviews for a few years, but it is still pretty easy as a reviewer to guess where the papers are coming from. We've done it without the blinding. Either way are complaints of bias.

The best suggestion I heard last this was discussed was to allow any author to put *any* name on a paper submitted for review. Your own name, your competitor's name, anyone's name you like. The intended effect was to cause confusion and second-guessing on the part of reviewers. I liked it. They didn't adopt it.

-Kevin

In general, the science community tends to favor those who are known within the community. Reviewers many times receive papers that are from their colleagues. Oftentimes, reviewers can be suggested. It would be hard to remove someone's name, as the science and research itself tends reveal the identity of the author. The "sub-communities" within the larger scientific community are even smaller. People know each other. I think that even gender neutral names might not be "scientific enough." I would go with a letter and number system. When students sign up for class they receive grades by this type of system. My code for a course I'm taking now is G33. It has nothing to do with any name.

Finally, I would like to comment on a story I heard. There was an author (female) who submitted a paper to Science. The paper was rejected. She resubmitted with the name of a well known person in the field (male). This person agreed to be on the paper to help it get published. The resubmitted paper was published supposedly with the same content. Sometimes we need name recognition and others who have paved the road ahead to help us on top of the outstanding science that is done. Unfortunately, it seems that is way the science world and world in general works - based on the apprenticeship model, who you know and your reputation.

An editorial on double-blind peer review was just published in Nature, and the text is reprinted in their Peer-to-Peer blog. The editors are soliciting comments and suggestions, and so far the count is up to 43.

I like the idea of only using initials - it's simple to implement, and gets round the problem.

I'm not convinced about double-blinding, but I don't think it would be disastrous either. There's no consensus that any one method is best, so I've decided that the best approach is to have a variety of review methods, and let people pick the ones they find best. Of course, the down side is that this biases the whole process towards game theorists.

Bob

As another variation on the "known" theme, my own suspicion is that along with being well-known, location "matters". There were a number of letters to editors some time ago from a scientist who moved from a top European lab to his home country and his ability to publish in the top journals fell distinctly. Same person, different location; different perceptions, perhaps?

I would agree with many previous posters that the best reviewers will know who the authors of a paper are, whether names are included or not (and often, the authors can figure out who the reviewers are as well). It's a small, small world.

One thing hasn't really come up in this conversation is that most papers have several authors, often of mixed sex. When I (male) submit a paper with my adviser (female), and it is rejected, can I blame it on sexism?

anon:

"..several authors, often of mixed sex."

Presumably each author usually has just one sex ;-) Couldn't resist...

Of course, I'm assuming that human hermaphroditism isn't common... (Nothing against hermaprodites, btw.)

In my field (physics) it is customary to use only initials--only rarely will an author use a first name. That must be why women are so successful in physics.

I can't imagine anyone reading this far down, but this article (http://arstechnica.com/journals/science.ars/2008/02/17/aaas-ethics-in-s…) summarises an AAAS session on science publishing which includes the observation (there are two links included in the original text, to see them read the original):

The topic of double-blind peer review was also raised; a number of studies have suggested that keeping reviewers in the dark about the authors and institutions that produced a paper would increase the diversity of both authors and affiliations appearing in major journals. These issues were covered recently at Scientific American, which was responding to a decision by Nature not to engage in double-blind review. The issue is interesting, but the discussion was not, as most seemed to favor going the double-blind route; Marusic suggested that 70 percent of authors also favor it.