This App Will Tell You if Your Job Posting is Sexist
If we want to #changetheratio, let’s start with how we talk to our ideal job candidates.
To figure out why there are only men on this panel at a 2013 tech summit, we might want to start by looking at the way the jobs they applied for were described. Image via Flickr user William Murphy (cc).
Programmer Kat Matfield has designed a free tool called the Gender Decoder to analyze job advertisements for subtle gender bias. Matfield says she was motivated by her “frustration at the gap between theory on bias and practical steps people can take to counter it.”
Researchers have long examined gender bias, though such studies tend to describe the problem, not the solution.
GUYS we did it! We traveled back in time to 1950! Oh... wait, no, that's just a sexist job desc. @Vestra_Inet https://t.co/SBXQzvprMv— Lauren Souch (@Lauren Souch) 1448295378
Matfield wanted to design a tool to help companies prevent bias in their hiring process, right down to the language used to describe job duties. On Matfield’s web page, users paste job ads into a text box, which tallies masculine- and feminine-coded words. Although English doesn’t utilize formally gendered words like “el secretario” (“male secretary” in Spanish) or “patronne” (“female boss” in French), our words do have subtle gender implications. For example, we tend to use “bossy” to describe women and “arrogant” to describe men.
Matfield was inspired to create her app by a 2011 study on gender bias in job ads, in which researchers changed the wording of ads to include more feminine or masculine words, then measured responses. Women found the ads with masculine language less appealing, and felt like they didn’t “belong” in those occupations. Men were slightly less drawn to job ads with feminine language, though it’s interesting to note that feminine language didn’t necessarily influence whether men felt as if they belonged in those roles or not. The research suggests that by describing the same job with feminine or gender-neutral language, companies will attract a more gender-balanced applicant pool.
Matfield’s gender decoder identifies masculine and feminine words based on word lists from that specific study. Companies are able to analyze and adjust the language of their job ads to appeal across gender lines. Since releasing her app, Matfield says a couple of organizations have asked her to build custom versions for their unique job sectors. She’s also discovered a complementary set of Chrome extensions, called Unbias, that allow users to hide pictures and names from social and professional networks, giving hiring teams a chance to vet potential employees in a gender-blind manner.
Ultimately, Matfield’s tool reveals just how deep sexism runs in our language. The words “challenge” and “analyze” read as masculine; the words “support” and “empathy” read as feminine. So maybe it’s not simply about changing our job descriptions to read more gender-neutral. We also need to change our culture so that words like “adventurous,” “understanding,” “opinionated,” and “gentle” are suitable descriptors for both men and women.
It’ll be a slow, culture-wide shift, but tools like the Gender Decoder and Unbias should serve as a strategic first step in inviting more women to apply for positions historically held by men—and vice versa. And there may already be hope for a more generous, less divisive workplace in the future. If you, like GOOD, want to explicitly avoid hiring divas, fret not: A quick check using the tool revealed the term to be gender-neutral (though perhaps that’s because it wasn’t a common enough term in job descriptions for the 2011 study to include).