Criticism often levelled at AI in decision making in general is that it is difficult to understand how the decisions are being made. It is some kind of black box in which a question is fed in, the algorithm is developed and somehow a decision is made, and the black box spits out a result?
How can this be used in an interview session when transparency, openness and defensibility of the decision are needed?
We have invested significant resources in this area, eliminating a range of inevitable bias that we, as humans, bring to an interview situation. Our video assessment platform asks interviewees to record their answers to client-defined questions. The responses are recorded – but only the words said. We are not looking at visual cues and scoring them, as there is no evidence that the technology is suitably advanced to make this a valid approach. We focus on the spoken words only, using AI to search for the positive and negative indicators, as any interviewer would do in an interview. We have developed a ‘glass box approach’ in which it is very clear how the algorithm is calculated.
Case example: making interviewing objective at a global travel company
A worldwide travel firm understood the need to update its hiring processes. Interviews were long and in-depth but not linked to the company’s core values. Assessments were used but they had little connection to the business and were used inconsistently. Work samples lacked scoring objectivity.
It introduced video interviewing and a more valid, objective and evidence-based assessment process to identify the best candidates and optimize recruitment team's and the candidates' time.
“Admittedly, the idea of a video assessment did make me nervous, but actually it was fine. The questions were well worded and gave me plenty of time to prepare my answer before recording it. I could also complete the interview at a time of my choosing.”Applicant to a global travel company