Wednesday, September 29, 2010

The Kuranta-Bothata Study

For many years I've felt that authors, referees, and editors--that's everyone, by the way--involved with scientific publishing have a bias against "weird" (third-world) names. That you are more likely to get published and/or to be cited if your name is Pratt (rather than Onwubulu). Frank Sinatra used to complain that his problem was having the "Sinatra" name, and that a name like "Pratt" would make life smoother for him. People I've met in the North say that this is ridiculous and preposterous, people I've met in the South say it's not only obvious, but just part of life.

So why not design an experiment? Here's what we did: We got a scientific paper with 6000+ words. The paper had no authors or addresses, but there was a manipulation: on the first page, there was mention that the work had been funded by an African-sounding agency or an European-sounding one. "This work has been generously supported by grant 26/110.790/2009 from the High Scientific Council of the Kuranta-Bothata Province." The imaginary European "province" was from the Dutch: "Welgesteld-Tijdschrift".

So there. Two words in a paper with 6.142 words. Can these two words change the perception of the entire paper?

Yes it can. While we're still crunching the statistics (we've just finished collecting the data from potential evaluators yesterday); the average scores for questions like "is this paper publishable", "would you cite it?", and "Should it receive additional funding?" are definitely very different.

Besides the obvious fact that one could criticize the external validity of any study of this sort, it seems, rather sadly, that people on the South are right about this issue and prejudice rules the land.