The potential for bias is as prevalent now as it has ever been.
From old school networks to outright discrimination, the potential for bias in the recruitment process, subconscious or otherwise, is as prevalent now as it has ever been. Legislation and company policies do, of course, play an important role in levelling the playing field, but they only go so far in achieving complete equality of opportunity. How do organisations ensure those with the desired experience, qualifications and skills aren’t missing out because of biased or discriminatory hiring practices?
Here at Elbo, we believe every candidate should have the right to judgement without bias and to be selected on merit alone.
We don’t believe in any kind of discrimination which is why our ethical algorithms are completely agnostic when it comes to matching candidates with their perfect role. Technology has an important role to play as we all work to remove bias from the hiring process, and here’s how:
Beyond the CV.
An important step in overcoming bias in recruitment is to think beyond traditional CVs. A CV is essentially a sales pitch – often tailored to a specific job spec – and therefore risks containing inaccuracies when describing an individual’s skills and experience. If the starting point is skewed and inaccurate, the outcome, namely the decision to hire someone, could well be the same. Put that CV in the hands of a recruitment agent with often limited understanding of the role they are matching, especially when it comes to technology roles, and it’s easy to see how bias can creep into the process.
Employers need, and should expect, a lot more information on a prospective employee or contractor. This is why Elbo’s technology collects multiple data points and draws shortlists using complex and ethical data models, intelligent weightings and sophisticated algorithms. Decisions are not left to potentially biased human judgement, but by matching roles to skills, psychometric testing, preferences and character traits.
Discount sensitive data.
Information on gender, ethnicity, country of birth or age, which could potentially be the cause of discrimination against a candidate, should not be captured or provided as part of an application to a prospective employer. Our platform not only discounts the need for this data, but candidates have the option to turn on incognito mode which hides their name and profile picture, further reducing the likelihood of human bias when applying.
Put it this way, we need to find ways of capturing and presenting data to potential employers which assumes biases will come into play when selecting a shortlist. How, for example, should an individual be judged if they don’t have a degree? Will someone with the right skills and experience be discounted on account of their age under the assumption they won’t fit culturally with a younger team? If someone has English as a second language, do we then make the judgement they won’t be a good communicator? Add to this that we traditionally look at volume over value when it comes to recruitment and it’s easy to see how exceptionally talented individuals can get lost in the mix. We need to look beyond these outdated criteria if we want to find the best talent.
Meritocracy is king.
At Elbo, we believe in the individual. We believe every candidate should be judged on earned skills, capabilities, experience and whether their personal traits and preferences are a good fit for the organisation – none of which are dictated by characteristics such as gender or race. When we take defined personal categories out of the selection process, and enable individuals to be judged on merit alone, we are one step closer to taking bias out of recruitment.
The AI myth.
Don’t be taken in by the AI hype. There’s a common misconception that the use of AI and machine learning is an effective way to eliminate human error and biased decision making. We only need to look at the work Google is doing to protect its own algorithms from bias AI to know this is not the case.
In recruitment, the use of AI and machine learning is still very new and often used as a marketing term for automating key word matching. Due to relatively small volumes of data and the overly basic, human teaching of learning models using the unreliable trio of agents, CVs and job specs, human biases inevitably creep in when creating shortlists. Similarly, where AI is used, the data loop is often closed too early – perhaps at decision to interview or hire – whereas Elbo is set up to capture structured data from the start of the application through to the end of the period of employment so has more extensive and richer data to draw from.