The analysis of PredictiveHire‘s model, which uses text-based communication to interview candidates, has been published in peer-reviewed journal IEEE Access.
The researchers used data from more than 46,000 job applicants who completed an online chat interview and a questionnaire based on the six-factor HEXACO personality model. The HEXACO traits are honesty-humility, emotionality, extraversion, agreeableness (versus anger), conscientiousness, and openness to experience.
Personality models such as the Big Five and HEXACO are based on the ‘lexical hypothesis’ that personality characteristics are encoded in language, showing the foundational impact of language in defining identifiable personality traits, the researchers say.
After the applicants’ personality traits were assessed they were asked to provide feedback on the accuracy of how they were described, and the researchers found 87.8% of the participants agreed with the description given for each of the six traits.
To avoid focusing on hypothetical scenarios that create potential for candidates to give similar answers to others, the interviews are oriented towards behavioural, not situational questions, PredictiveHire CEO Barbara Hyman tells Shortlist.
Candidates can likely work out what trait is being assessed by each question, but they can’t “game” their responses with pre-rehearsed scenarios, she says.
Examples of the questions used in the interviews include:
Candidates respond to the assessment questions with a noticeable sense of intimacy and authenticity, even including emojis in their answers – “the same way they would respond to a friend”, says Hyman.
She adds that a text-based approach leaves less room for recruiter partiality compared to reviewing CVs, traditional psychometric assessments, and video interviews (which she describes as a “petri dish for bias”).
Predicting personality using answers to open-ended interview questions, IEEE, June 2020
Source: Shortlist.net.au | Wednesday 15 July 2020 9:21am
Get our insights newsletter to stay in the loop on how we are evolving PredictiveHire