Nearly everyone agrees that science has a gender problem. But the size of the gap depends on the area of science. Now, a study of nearly 1 million engineering paper co-authorships puts hard numbers on the problem in this male-dominated scientific field, and finds a paradoxical trend: Female engineers are publishing in slightly more prestigious journals on average than their male colleagues, but their work is getting less attention.
Across the STEM fields—science, technology, engineering, and mathematics—there are gender disparities in tenured faculty positions, publication rates, and patents awarded. What's unclear is exactly how these differences arise. The new study makes use of bibliometrics, or the statistical analysis of publication patterns in data culled from vast numbers of scientific articles, authors, and citations. Its lead author, Gita Ghiasi, is wrapping up an engineering Ph.D. at Concordia University in Montreal, Canada.
The researchers bought the data from Web of Science, the massive—if somewhat flawed—database of scholarly publications maintained by Thomson Reuters. Filtering for engineering journals published between 2008 and 2013, the team harvested 679,338 articles with nearly 1 million co-authors. To assign gender, the researchers used databases of male and female first names specific to the country of the author's affiliation. To distinguish among the various engineering subfields—aerospace technology, chemical engineering, civil engineering, and so forth—the team used the field designated by the National Science Foundation for each journal. Finally, to rank the journals by prestige, they used each publication’s Impact Factor, the long dominant—if increasingly unpopular—metric provided by Thomson Reuters. Then the team built a network map of collaboration across this 5-year slice of engineering publications.
The results show that women made up just 20% of the authors on the engineering papers, compared with about 30% across all scientific disciplines. But there is also a bit of good news: Engineering papers with a woman as lead author were published in more prestigious journals, on average, than male-led studies, the team reports 4 January in PLOS ONE. The difference in score amounted to a 2% bump in Impact Factor. But in spite of publishing in more prestigious titles, women's papers were cited about 3% less frequently. Those differences are tiny—which is part of the good news—but with hundreds of thousands of data points, they can't be waved away as a statistical fluke.
The team doesn’t commit to a particular explanation for the gap. But Ghiasi and her colleagues do offer some ideas on how it might be reduced. They found that women play a more marginal role in the collaboration networks identified by their co-authorship analysis, and it's not just men to blame: Women also co-authored multiple papers with female colleagues less often than chance alone would dictate. "Women engineers are complying with the male-dominant engineering scientific system instead of changing its structure," they conclude. They say women could close the gap if they were to collaborate with each other as often as they do with men.
The study is "very important because of the large sample size," says Joyce Benenson, a psychologist at Emerson College in Boston who found a similar pattern of gender bias in scientific collaboration. Her own research echoes the patterns of collaboration that Ghiasi’s team found within different areas of engineering. "When there are fewer individuals in a field, women play more central roles," she says. But overall, she is happy to see that the gender difference in prestige and citation is small. "Women are consistently doing really well in engineering across many subfields. It’s really a nice finding, and very positive."
"I definitely collaborate as much as possible," says Katy Huff, a nuclear engineer at the University of California, Berkeley. "I find better science results from the constant feedback among peers inherent in collaborations." She acknowledges that the engineering workplace poses an extra hurdle. But she's not convinced that the problem can be solved by engineers themselves. "Societal norms at large, not only the societal norms of the engineering research workplace, must change."