Technical University of Denmark: New technology minimizes recruitment bias
Job seekers will often encounter the wording “all interested parties regardless of age, gender, race, religion, and ethnicity are encouraged to apply” when reading companies’ job advertisements.
About a year ago, DTU went a step further and introduced the addition “to avoid any bias in the recruitment of new employees, we have introduced blind recruitment”. This was done in a single job posting in an attempt to minimize unconscious bias and ensure that all job applicants—regardless of their background—have the same conditions for being considered for a position.
New tool anonymizes applicants
At DTU, blind recruitment is tested using new technology developed by Sara Shafiee, who is co-founder of the start-up DivERS and a researcher at DTU Construct.
“We’ve developed a tool that anonymizes and automates the initial part of the recruitment, called shortlisting. With our tool, the hiring manager is not given any information about the applicant’s name, age, or appearance. All that remains is to find out how the applicant matches the specific job criteria. This means the requirements made for the applicant’s experience, skills, and education,” says Sara Shafiee about her new invention.
Normally, blind recruitment involves an extensive and time-consuming process. This is due to the manual screening, where you screen all applications and cross out the passages that highlight the applicant’s identity markers. But with the new tool, this may be a thing of the past.
“The tool automates the process by collecting the necessary data about the candidates and analysing how well they meet the requirements in the job description. It can make a list of the top five candidates in just a few seconds,” says Sara Shafiee.
As part of the online application process, the applicant must answer some screening questions that the tool generates based on the job description.
“The answers trigger different points, which gives the individual applicant an overall score. The higher the score, the greater the chance of being considered for the position.”
Sara Shafiee is one of the few female researchers at DTU Construct, and it is close to her heart to find solutions to the challenge of a lack of diversity in engineering.
“With our blind recruitment tool, I hope that we can speed up the development by minimizing bias in the selection of candidates for job interviews,” says Sara Shafiee.
DTU Professor tests tool
The experiment with blind recruitment was conducted in collaboration with Professor Vincenzo Eposito from the DTU Energy. Last year, he was in charge of filling a PhD position for a new research project on new materials for batteries.
On expiry of the application deadline, 80 applications had been received with accompanying CVs and appendices, all of which were screened manually and assessed equally thoroughly by Vincenzo Eposito.
“I usually spend several weeks reading CVs and applications out respect for the applicants and to treat them fairly. I’m a foreigner myself, so I know how difficult it can be to get through the eye of the needle,” says Vincenzo Esposito, who participates in recruiting new employees at least four times a year.
In all its simplicity, the experiment was about comparing the Professor’s selected candidates with those of the tool. The result showed a large coincidence between the candidates ranked highest by the Professor and by the tool. The second choice of the tool was consistent with the candidate chosen by Vincenzo Esposito as the best qualified candidate, who ended up being hired.
Breaking with habitual thinking
Vincenzo Eposito will not go as far as to say that the experiment has changed his behaviour, but he sees the new technology as a useful tool in breaking with conventional thinking.
Not least, the experiment made him realize how much is at stake when he selects candidates for job interviews.
“It’s one of those situations where you don’t want to risk too much. You want to be sure of finding someone who is at the right level in relation to your research and who can deliver results. It’s therefore completely understandable if you choose to follow your own gut feeling and do things as usual,” says Vincenzo Eposito and continues:
“If the use of blind recruitment and a tool like this one can help put together a more diverse team, then I’m happy to be able to contribute to this development.”
Blind recruitment is still an experiment that is being further developed and adapted by DTU. DTU is also working with the wording of job postings so that more candidates choose to apply.