Select Page

Measuring leadership potential has become a norm in many industries, and for a good reason. The data can help you determine which employees have leadership potential and understand your company’s individual management culture.

The most common method of collecting this data is via leadership surveys, but recent studies have started revealing troubling levels of bias within this process. According to independent evaluations of this data, tall white men are generally judged more favorably than their colleagues who are short, women, people of color, or who have differing values to the people conducting the surveys. Surveys suffering from these biases do very little to measure actual leadership potential – rather, they measure who is perceived as a good leader by the people evaluating them. Here are some strategies to reduce bias and get better leadership data.

  • Determine Your Evaluator’s Ideal Leader

Knowing what an ideal leader looks like to your evaluation team before they start rating real candidates can really help you achieve more accurate data from these surveys. Asking people to name the qualities of an ideal leader can make them more aware of the filters through which they perceive people and can help them temper their bias.

  • Require Specific, Qualitative Examples

When someone is rating someone else’s behavior, you should always ask them to provide specific examples of the behavior they’re noticing. Asking for concrete examples makes them reflect more deeply on their answers and calibrate their ability to provide an accurate rating.

Providing these examples can be difficult, especially in the case of sensitive information, so taking steps to reassure your evaluators that their responses will be confidential and used appropriately can also help reduce bias.

  • Make Them Slow Down

Biases are most likely to appear when decisions are made quickly, so slowing the process down can yield more accurate, considered results. Designing ways to make people stop and think about their responses can reduce biased responses by up to 75% and make the results you gather more clear and indicative of what you’re intending to measure.