How Data Scientists Are Working To Improve Equitable Hiring Practices

Data scientists can help improve and implement equitable hiring practices by detecting and removing bias in job postings, candidate screening, tests, and interviews.

75% of organizations surveyed by law firm Baker MacKenzie said they use technology including artificial intelligence for hiring and other core HR activities. As employers increasingly digitize hiring and recruitment, data scientists are being tapped to ensure hiring tools implement equitable hiring practices. Let's take a closer look below.

Technology Enables Equitable Hiring Practices

Surprisingly, the survey reports that IT and tech people (instead of HR) are being tapped to manage technology-enabled hiring tools once they are in place. This statistic implies that tech talent, including data scientists, is shaping the future of hiring in companies. Indeed, 95% of surveyed companies report they have a person or team responsible for ensuring AI bias is evaluated in digital tools before hiring systems are bought and installed.

Below are some ways data scientists are working to improve equitable hiring practices in the hiring process. 

Train AI To Remove Biased Keywords

Data scientists can use artificial intelligence to ascertain the presence of bias in job postings and descriptions. Intelligent tools that review job posts are programmed to highlight keywords such as:

  • Gender-biased terms: These include gender-stereotyped descriptions like: “driven,” “competitive,” “expert,” “analytical,” “caring,” “compassionate,” etc.
  • Ageism: Verbiage like “young,” with specific years of experience, and exclusive age brackets such as 25-30 years old.
  • Marital status: Remove terms like “single,” “unmarried,” “preferably single,” and “married.”
  • Physical appearance: Some common biases could sound like “must be 5'5 in height,” “must have a good physical appearance,” “good looking,” and “attractive.”

In the same way, data scientists use AI to review the biased tone and phrasing in job postings. Consider the following examples:

  • Male-oriented: “Our team is searching for an active, strong candidate who thrives in a competitive atmosphere.”
  • Neutral: “Our team is looking for excellent individuals who want to grow in a community of highly motivated professionals.” 
  • Female-oriented: “Our team is looking for caring individuals who serve and nurture customers.”
  • Neutral: “Our team is searching for frontline professionals who provide excellent customer service.” 

By detecting and removing biased keywords, data scientists help implement equitable hiring practices at the start of the process, so more candidates feel welcome to apply. 

Match Candidates With Positions

Data scientists can write algorithms to match candidates with suitable positions based on their skills and expertise. They can also develop recommendation systems that invite candidates to apply for roles they may not have considered initially due to perceived bias or stereotypes in the hiring process. 

For example, an AI-enabled system is trained to learn about candidates’ interests and areas of expertise. AI then includes job search results that would not usually come up with the users’ keyword search, such as executive coaching for those with teaching experience, interior design for graphic artists, etc.

Focus on Skill 

Putting the spotlight on skill is one of the best ways to reduce bias. There are several ways that data scientists emphasize skill in equitable hiring practices:

Job posting: Data scientists can write algorithms that highlight skill-related terms instead of personal attributes, length of experience, education, location, and other non-skill factors.

Candidate screening: Technology-enabled tools can be programmed to find skill-related words in candidates’ resumes. Tools can also be used to automate background checks on related professional networks.

Assessment: Instead of assigning company staff to give tests to candidates, data scientists can develop or use digital hiring platforms that automate the administration of remote skill tests. Technology also plays a crucial part in automating scoring, so evaluation stays objective, impersonal, and focused on skill. 

Interviews: Data scientists can employ natural language processing in video interviews to check candidate speech content for key skills and competencies. This information can then be passed on to human decision-makers for more insightful candidate evaluation. 

Success prediction: Data scientists can analyze historical data regarding hiring decisions, including skill-based interviews that led to a good hire. This information can then be used to predict candidate success in the job role. 

Screen Datasets And Model Outcomes For Bias

While AI and machine learning algorithms can be programmed to reduce bias in hiring, these tools are still subject to the basic principle of garbage in, garbage out. If the datasets used to program hiring tools contain unconscious bias from the people who made them, the end result will still be tilted toward a favored group. Data scientists help keep hiring tools neutral by checking:

  • Are data sets dominated by specific groups?
  • Are data sets restricted to certain characteristics only? 
  • Are predictors in models unusually correlated with non-skill factors such as age, race, or gender?

Data scientists can also check if model outcomes are delivering results aligned with a company’s equitable hiring practices. By regularly monitoring, analyzing, and improving algorithm design and models, data scientists ensure that technology-enabled tools stay neutral and inclusive. 

In summary, data scientists can help improve and implement equitable hiring practices across the hiring process in various ways. These include training AI to remove biased keywords, using recommendation systems to match candidates with positions, highlighting a focus on skills during assessment and interviews, and regularly analyzing datasets/models for bias.

Improve Equitable Hiring Practices with Filtered.ai

Startups as well as Fortune 500 companies have used Filtered to implement equitable hiring practices in their workplaces such as:

  • Improve inclusion hiring
  • Increase diversity
  • Make promotions based on performance
  • Conduct performance reviews more meritocratic
  • Interview technical talent from developers to data scientists based on skill
  • And more 

Filtered is a leader in skills-based, data-driven recruiting technology. Our end-to-end technical hiring platform enables you to spend time reviewing only the most qualified candidates, putting skills and aptitude at the forefront of your decisions. We’ll help you automate hiring while applying objective, data-driven techniques to consistently and confidently select the right candidates. To get started, contact our team today or register for a FREE demo.