Mitigating gender bias in structured interview responses

Language differs with gender. If we’re not careful, language can lead to gender bias.

This scientific paper details a new experiment on the structured interview, and how it can account for, and minimize, the gender bias that accompanies the traditional interview.

Data for this experiment comes from 633,413 candidates (52.89% female) who applied to two large retail organizations in the UK and Australia. Candidates participated in an online chat-based structured interview where they answered 5-7 open-ended questions related to past behaviour and situational judgment on the Sapia’s Chat interview platform.

Our work shows that language data from structured interviews can be used to score candidates without inducing gender biases by using a set of well-defined scoring dimensions (features) despite the raw responses carrying higher levels of gender information.