A new Harvard Business Review article just came out featuring exciting research by University of Minnesota alumni and faculty. In Hiring, Algorithms Beat Instinct says that, “humans are very good at specifying what’s needed for a position and eliciting information from candidates- but they’re very bad at weighing the results”. Empirical evidence supports that a simple equation can outperform human decisions by at least 25%! This accuracy discrepancy can be attributed to people’s tendency to be easily distracted by things that may be only marginally relevant. For example, if you found an applicant that went to your alma mater you are likely very excited and make sure to give that applicant a second look, but is that applicant really more qualified than one who went to a comparable institution?
There is strong resistance to the idea of purely algorithmic selection and the authors do not advocate for this extreme. Rather, they encourage human resource managers to use an algorithmic system based on a large number of data points to initially narrow the field. Once you have pruned your selection pool using research supported selection tools (i.e. AAI’s Work Behaviors Inventory or Applied Reasoning Test) managers are encouraged to use other personal selection tactics that involve intuition in their final decision making process.