Measuring women’s participation in journalism once meant sitting down with a stack of newspapers and counting bylines by hand. That’s no longer the case, thanks to computer programs that use big data to examine gender biases in sourcing, story placement and even retweets.
The results so far are grim, with women remaining chronically underrepresented in many aspects of news. But the creators of the new tools hope the information they collect will help journalists assess their habits, and perhaps change them.
Each piece of software works a bit differently, but the basic concepts are similar: Computers comb through online articles and compare the names of authors and sources with databases that determine if those names are likely male or female. The results aren’t perfect, but they can reveal broad patterns.
“They might not be as accurate as thousands of people looking over articles by hand over a period of five years, but they can give you a rough check before you hit that publish button,” said Nathan Matias, a graduate student at MIT’s Media Lab and Center for Civic Media. Read more