Removing Ableist Language from Job Descriptions with the Department of Labor Office of Disability Employment Policy
The Client
The US Department of Labor Office of Disability Employment Policy (ODEP) was created in 2001 to create and implement policies and changes to increase the number and quality of employment opportunities for people with disabilities. Over the past 20 years, ODEP has driven a wide range of policy changes that impact the employment of people with disabilities, including initiatives focused on transitioning youth with disabilities, workplace policies and practices, and employment supports such as accessible technology and transportation.
The Problem
What is ableist language?
Ableist language is language that is offensive to people with disability and can also refer to language that is derogatory, abusive or negative about disability. Ableism is the systemic exclusion and oppression of people with disability, often expressed and reinforced through language. [source]
Why is this tool important?
Ableist language in job descriptions can cause people with disabilities to feel excluded from jobs that they are qualified for. This typically occurs when a description references abilities or enduring attributes of an individual that are unnecessary for the job or for which accommodations can be proactively offered instead of focusing on developed skills that can be acquired to succeed in the role.
The Solution
Description of the project
The Department of Labor ODEP and xD Census built out an NLP-powered tool to identify ableist language in job descriptions. This tool assists employers in creating awareness and actionable insights to ensure companies are being inclusive to people with disabilities. It checks a given job description against an ableist language lexicon and recommends alternative words or phrases that are more fair and inclusive.
Main Features
After inputting the job description in the text field of the home page, the web app processes the text through the NLP algorithm and brings the user to the results page. Functionalities behind this application involve
- Counting the number of instances of ableist language, including duplicates, meaning that if the ableist term is used twice in the job description, the app will identify and count both instances.
- Locating the ableist terms within the job description. The page displays the entirety of the user’s input with the ableist terms highlighted. Mousing over these highlighted words/phrases will give the user the ability to click on the word to see further details.
- For each ableist term, there is a mapping to possible alternative verbs. Once the user clicks on the word, the information box on the side of the page (or bottom for mobile view) will display the alternative suggestions that are more inclusive as well as an example of how to use one of the alternatives.
See a video walkthrough of the web navigation here.
Reflections
Our team tackled the challenge of building a government website from scratch, incorporating elements from the industry pipeline and keeping in mind accessibility and mindful design. Making sure that no small detail was looked over was something the team strived to do at each step, and we worked toward a final product that was a result of us collectively learning a lot about how websites are deployed and tested in the real world.
“I personally learned that good design is flexible design; incorporating styles that were friendly across multiple devices and for many different users reminded me that we should be human-centered in all the products we make. It can be easy to underestimate just how important intuitive flows are for a website, and making a government site easier to use for all is a good change from the status quo.” -Kevin Tan (Software Engineer/UX Designer)
“Working on this project has taught me a lot about how soft skills, such as communication, conflict resolution, and organization are just as important as technical skills such as writing code and creating designs. Knowing how pieces of the project fit together, having a long-term vision for the code development process, and being prepared for the worst case scenarios have been essential in succeeding in this project.” -Jamie Lu (Software Engineer)
The Team
Kevin Tan
Kevin (he/him) is a sophomore at Harvard College studying Chemical & Physical Biology and Computer Science. He hails from the sunny Arcadia, California and is a proud San Gabriel Valley native. He is driven by his desire to unite community power, science, and mindful design to build more equitable infrastructure and beautiful spaces.
Jamie Lu
Jamie (she/her) is a sophomore from the Bay Area, California studying Computer Science. She is passionate about the intersection of technology and business innovation to heal pressing gaps and challenges in people’s lives.
Zad Chin
Zad (she/her) is a sophomore from Malaysia at Harvard College studying Computer Science and Economics. She is interested in using technology and entrepreneurship in contributing to social good, particularly in the sector of health sciences.
Acknowledgements
We would like to thank Harvard Computer Society Tech for Social Good and the Department of Labor ODEP for connecting us with this wonderful opportunity to further our skills by developing this web application. We also want to thank Maria Patterson and Diana Lam for their support throughout the project and for being such collaborative clients. Their insights on the best industry practices really helped our team grow and flourish, and we are incredibly grateful for this partnership.