|Attracting and retaining talent is difficult enough, but during a skilled labor shortage, it’s even tougher / Photo by: iJeab via Shutterstock|
Attracting and retaining talent is difficult enough, but during a skilled labor shortage, it’s even tougher. A recent survey conducted by CareerBuilder, an online employment website, found that 44% of businesses are looking to hire full-time employees and 51/% are planning to hire temporary employees. However, 45% of them are unable to fill much-needed positions due to the dearth of qualified talent.
A 2018 report also showed that talent shortage was one of the biggest hurdles faced by recruiters the previous year. It was reported that 72.8% of employers were struggling to find relevant candidates and 42% were worried they won’t be able to find the talent they need. This meant that recruiters should come up with a different hiring strategy. Fortunately, artificial intelligence and machine learning can help firms recruit and retain employees.
AI algorithms can look into a candidate’s job application and decide which applicants show the most promise. Experts stated that these algorithms take out all subjective bias from human recruiters during the early stages of the recruitment process. After the candidate has passed the initial screening, humans would perform a more in-depth assessment of them. By this stage, the chance of losing out on a great candidate to some form of recruiter bias has been drastically reduced.
According to Bold Business, an online site that offers digital marketing, business technology solutions, employee training development, and business process management services, recruitment algorithms can be used not only for resume review but also for cover job advertising, resume ranking, and late-stage employee selection. This can help in making processes faster and accurate.
Does Using Algorithms Prevent Bias?
Just like other AI technologies or tools, recruitment algorithms have also some issues. In theory, the use of AI and machine learning hiring systems should eliminate human biases that may exist during the hiring process. By using objective data, the bias that HR personnel has when recruiting, interviewing, assessing, and making hiring decisions would be reduced or eliminated. The end result should, therefore, be a more fair and accurate decision about hires. While this is great news for companies, research showed that the systems have biases of their own.
A recent example of this is Amazon’s experiment into an AI-based recruiting system, which was removed after it started penalizing resumes that included the term “women’s” or names of women’s colleges. The system also taught itself to prefer resumes of male candidates over those of female candidates.
Unfortunately, accurately determining if recruitment algorithms are fair or not is not an easy task. Researchers from Cornell University discovered that only a few companies offer concrete information about how they validate their assessments or disclose specifics on how they mitigate algorithmic bias. While many of them acknowledge the impact of hiring algorithms and are taking steps to address bias and discrimination, they provide little to no information about their assessments.
|Just like other AI technologies or tools, recruitment algorithms have also some issues. In theory, the use of AI and machine learning hiring systems should eliminate human biases that may exist during the hiring process / Photo by: Vintage Tone via Shutterstock|
"Plenty of vendors make no mention of efforts to combat bias, which is particularly worrying since either they're not thinking about it at all or they're not being transparent about their practices," author Manish Raghavan, a doctoral student in computer science, said. The study "Mitigating Bias in Algorithmic Employment Screening: Evaluating Claims and Practices” also recognized that algorithms have the potential to contribute to a more equitable society despite their flaws.
According to RecruitingDaily, a company that deals in recruitment, human resources, and talent acquisition globally, the bias in recruitment algorithms happens first during the job advertisement process. Most of the time, job boards borrow the ad distribution logic from the real-world data where there are existing biases. As a result, job ads would only be targeting selected candidates. For instance, if the real-world data reflects the pattern of black candidates applying for low-paying jobs, the same pattern will be used to distribute ads.
Removing Bias From Recruitment Algorithms
While recruitment algorithms have great impacts on HR processes, researchers question whether they are reliable and valid. "Just because your decisions involve data doesn't mean they're more objective. The idea you would be able to infer personality traits that are genuinely relevant to hiring decisions from video interviews seems suspect to me,” said Solon Barocas, an assistant professor of information science at Cornell University.
Algorithms only show bias because they were trained to learn that. However, when they are trained in a fair, objective, consistent, data-driven, and inclusive approach, the existing bias would be eliminated. One of the ways to prevent bias on these algorithms is by building a model using data from representative samples. This means a model should be designed using data from candidates from all demographics in fair and representative proportions. It should be built with a 50:50 ratio.
According to LaunchPad Recruits, a job interview platform, companies should also make sure that only the relevant aspects of the job are assessed during the hiring process. This will not only prevent bias but also avoid inconsistent decision-making. This can be done by conducting a job analysis and respectively designing questions and review criteria that reflect those relevant aspects.
At the same time, it’s also important to hire a diverse team. This not only includes people from various backgrounds but also creatives, linguists, sociologists, and passionate people from other walks of life. Companies should also monitor AI outcomes of the hiring algorithms at all stages. This will help in understanding the scope, depth, and frequency of retraining the algorithms.
|While recruitment algorithms have great impacts on HR processes, researchers question whether they are reliable and valid / Photo by: fizkes via Shutterstock|