Bohnet, I., van Geen, A. & Bazerman, M. H.
Cambridge, MA: Harvard Business School
Primary Research
Area of focus: 
Unconscious Bias

This experimental research study tests a proposed new intervention to overcome gender biases in human resources processes. Employers hiring new staff typically use a joint evaluation process, comparing the expected performance of all possible recruits, whereas when allocating promotions and job assignments they more typically use single evaluation of an individual staff member's competencies. The authors argue that in separate evaluation the lack of comparison information invokes the evaluator's intuitive "fast-thinking", which can be prone to gender biases once the gender of the applicant becomes known.

In a controlled experiment simulating a performance and remuneration system similar to a corporate workplace, subjects in one condition were asked to decide whether to hire a given candidate based on their past performance in single evaluation, and subjects in another condition were given a candidate of each gender to choose between. The research found a significant gender bias in the single evaluation condition, but none in the joint evaluation condition where past performance instead became the significant point of variance.