How you can eliminate recruitment bias using predictive analytics
While employing people of different race and gender goes a long way towards improving organizational performance, it’s just a start. Diversity goes far beyond race, color, creed or gender. Diversity means just that – diverse characteristics, skill sets, experiences and personal attributes too. People who think differently approach problems in different ways. Developing a high performing, innovative and resilient workforce requires people to be open to divergent thinking and ideas at all levels in all ways. However, the brain is a social organ and human nature dictates that we prefer people who are like us. Hiring people in our own image only perpetuates our own limitations and stifles a business’s progress. For example, you might have a business full of extremely clever analysts who are not natural communicators, if they are unable to communicate their findings in a compelling way, then no-one will be interested in their data, hence you would be wise to hire people with communications skills too. Businesses therefore need to scrutinize their workforce not only to decipher what skills they are lacking in but also what attributes their highest performers hold and try to reflect those in future hiring decisions.
This is where predictive analytics, or big data, has a crucial role to play. It helps to eliminate human biases from the decision making process so its algorithms can identify candidates with non-traditional profiles but with skills or experiences that are similar enough to those of a high performer. By trawling through resumes, the technology can identify strong candidates by searching for specific skills and attributes regardless of background, ethnicity, social mobility or gender. It is looking for candidates that have the right abilities and experiences to deliver higher revenue and profit for the company and stay longer.
So how does predictive analytics really eliminate bias? It looks to the past for information in order to shine a light on current behavior and predict future behavior. Used correctly big data can help employers to identify and quantify any historic bias to eliminate it from future decision making; decipher what attributes make a great hire, mitigating the influence of discrimination creeping in; and understand which sources offer the most promising talent and therefore how to extend that to reflect social mobility concerns. In operational terms, big data also enables HR teams to provide stronger data and evidence to support hiring decisions. Plus it helps them to be more time efficient enabling them to concentrate on better engagement techniques to attract the best candidates ahead of the competition, and to hold on to the talent for longer.
However, big data is only as strong as the data it holds and the questions it is programmed to ask. It relies on the imperfect inputs and logic of the people who design it. Without deliberate care, the very system we rely on to eliminate bias can perpetuate the prejudices of prior decision makers or reflect widespread, persistent biases in our society. It is therefore imperative that HR and data officers work together to devise a predictive system that works best for the organization. This involves defining what you want your big data to do, for example making a recommendation to interview an applicant on certain pre-defined criteria. As well as building your algorithm you need to test it, by running it, analyzing it, restructuring the data if necessary and tweaking the algorithm until it’s right for your business objectives.
More importantly, make sure you have a strong pool of data and that you are using it in a meaningful way. Although most HR departments have business analytics, many still struggle to use it to its full potential. This is because their people data is fragmented across several systems: core HR, different payrolls, benefit carriers, talent management and financials. Also, many are only using reporting dashboards and analytics in an administrative and operational capacity when they should be using it to help support decision making, strategic planning, scenario planning, and business transformation.
Finally, when collecting any personal data, you have to be extremely careful that you comply with privacy and data protection laws. In Europe, companies must give notice to potential recruits of the purposes for which their data will be used and the regulation stresses the need for human judgement in the decision making process. It is not a case of “computer says yes” but rather an informed decision by humans based on the available data and the interpretation of that data. In the US, federal and state laws limit the types of information that an employer may lawfully request or consider in making employment related decisions, even if the information has been obtained lawfully.
Technology will never take the place of highly skilled HR professionals, but it can validate decisions and streamline operations in real time. Big data has the potential to accelerate hiring, drive innovations that reduce bias in recruitment decisions and stimulate businesses to be more transparent and democratic in their selection process.