About this tool

Why gendered, ageist, and ableist langauge matters in job descriptions

The type of language used in job descriptions makes a difference in who applies for a job or not. Women are often less likely to apply for higher paid roles. Likewise, those who are older may shy away from jobs that call for 'a go-getter attitude', or for someone to be 'dynamic'; and ableist terms might dissuade people from applying when they are more than capable of undertaking a role. This tool also now highlights when a job advert is missing a salary or salary range.

With ableist langauge, the US Department of Labor (USDL) provides a great tool on which this one is based. While code provided by the USDL might be open to criticism for substantiating positions with political motivations, I didn't feel when using it that this materially underminded any of the outcomes. It provided a template for how to develop this page to work across gender, age, and ableism. This way we can see where the issues arise and some options to make a change.

Why is this tool important?

Gendered, ageist, and ableist language in job descriptions can cause people to feel excluded from jobs that they are qualified for. By identifying these questionable pieces of language and suggesting alternatives, this tool will support more inclusive hiring practices. This tool will work across all industries and is not specific for publishing, though that's the point I began from.

How does it work?

The tool uses a natural language dependency parser to search for key terms that have been tagged as gendered, ageist, or ableist by the USDL and me.

Some of the terms come from my work on ableist language lexicon gender in job descriptions in publishing language. Likewise, the ageist terms comes from my research into this. I also used a variety of research to build up the lists.

Who built this?

The original language detection algorithm was built through a partnership between the Presidential Innovation Fellows, xD, and DOL ODEP. The web application was built by HCS T4SG, with support from PIF, xD, and DOL ODEP.

This particular version of the tool was built by me (Dr Miriam J Johnson) as a test to see how it works and if it's useful. The USDL tool is open sourced and you too can visit the links and get your own version - god speed in your version compatibility journey.

About the author

I'm Dr Miriam J Johnson, I'm an academic of publishing and marketing. I also am a co-founder of a AI-powered startup in the publishing space that makes personalised audiobooks. You can check out Dudley Editions here.

You can find out more about me here: here.