Understanding the science behind AI-based hiring assessments
In this podcast, Mike Hudy, chief science officer at Hirevue, says industrial-organizational psychology predicts success more accurately than old-school resumes and interviews.
Assessments are among the most valuable hiring tools, but they can be challenging to design and implement. As a result, organizations tend to use them for high-volume recruiting, where large numbers of people need to be hired in a short time for roles that can be encoded in a handful of reusable assessments.
Recruiters use assessments to determine whether a candidate has specific job skills, such as writing software, but they're often equally interested in identifying the cognitive capabilities and behavioral characteristics that are required for a job.
Assessments are usually digitized, automated and conducted on computers. They might be games that challenge problem-solving skills and test reaction times, simulations of work scenarios or timed tests of programming ability. Quizzes test knowledge in specialties like medicine, finance and design.
In this episode of Enterprise Apps Unpacked, Mike Hudy, chief science officer at Hirevue, explains how AI is making assessments easier to develop and deploy -- and, therefore, feasible for more specialized job openings, including executive positions and newly created roles.
Mike Hudy
Transforming soft science into rigorous assessments
Hirevue is a video interview pioneer that, in 2013, began using AI to screen candidates, in part by analyzing their responses to standardized questions asked in structured interviews. In February, the company launched Assessment Builder, which it said makes scientifically validated assessments accessible on a large scale -- even for low-volume roles -- while minimizing the traditional tradeoff between speed and quality.
The speed and scalability, unsurprisingly, come from AI, which analyzes job descriptions to help recruiters and hiring managers quickly create role-specific assessments. Users can choose how much AI to use, including when scoring assessments.
Industrial-organizational (IO) psychology is the science behind Hirevue's AI-assisted video interviews and assessments. In the podcast, Hudy, who earned a Ph.D. in IO Psychology from The Ohio State University, explained why it provides a rigorous approach to candidate evaluations.
"We've done over a thousand validation studies, which is basically testing it and looking at what we measure in assessment and at post-hire outcomes and establishing a statistical link between them," he said. "There is a very strong statistical relationship, if you do it right, between what you measure in an applicant and their ultimate success in the job."
Hudy said any fair evaluation of IO psychology as a science requires comparing it to the baseline alternative: relying on resumes and unstructured interviews conducted by recruiters and hiring managers. Research shows neither is a good predictor of success.
"We as industrial organizational psychologists, in a soft science, can improve the baseline of that tremendously," he said. "IO psychology does a pretty darn good job of getting it right, and advances in artificial intelligence and technology are helping us go even further down that road."
Other topics discussed in the podcast include the following:
- How Assessment Builder works.
- Steps Hirevue took to remove bias from the tool.
- Ways digital technology, including AI, is changing soft sciences such as IO psychology.
David Essex is an industry editor who creates in-depth content on enterprise applications, emerging technology and market trends for several Informa TechTarget websites.
