Scientific journals don’t track the gender of their authors. That made it that much trickier for University of Washington psychology professor Ione Fine and her colleagues to uncover a surprising fact: that women scientists are significantly under-represented among authors of studies published in top-tier journals.
But they found a workaround using a bit of computer code.
"What we did is a classic example of data mining," Fine told Living Lab Radio.
"We pull out the author’s name and we compare that to another database which also has many thousands of names, and it has a probability for each name of it being male or female."
They excluded any names that weren't clearly male or female, which added up to about 5 percent.
They categorized the rest and found that while 30 percent of highly competitive grants from the National Institutes of Health go to women, only 15 percent of articles in the journals Nature and Science credit women in the most prestigious locations (as first author or last author).
"So there’s a massive disconnect between these two measures of women’s contributions to science,” she said.
Now that her lab has been able to show this mismatch, Fine hopes that it will prompt some changes at the top scientific journals.
"Having a process that provides a service that is...probably quite discriminatory is a real problem, and these are multimillion dollar industries," she said. "They should take it very seriously.”