Engaging read on performance review description words and how subtle shades of language can vary by gender, using a large dataset from US Naval Academy (Go Army, Beat Navy). To caveat, the underlying paper (worth diving into...no naval pun intended) looks at peer reviews rather than formal officer evaluations. Still we should hold future leaders to a high standard as well.
After reading this, I looked back at my company command time and recalled that in rating platoon leaders I found descriptors such as "caring" and "enthusiastic" suitable for some male officers and "assertive" and "confident" for some female officers even as the chart here shows them often reversed by gender. I saw similar trends among my senior rater (battalion commander) comments.
Perhaps the Army or the Engineer branch is better able to reduce biases in assessing talent than some other organizations though neither is perfect, but it starts with leaders evaluating and mentoring in a way that draws out the best from everyone. This attitude and the positive role model example of female platoon leaders in my previous organization were in my mind key contributors to our success as one of the first in the Army to integrate female combat engineers who were just as equally part of the team as any of their peers.
Even when their performances are objectively equal, men and women are described differently.
Reinventing Software Development
NO! 6 isn't a magical number. The general thrust of the article is that FEWER data points results in less bias because we associate genius and exceptionality with males [and 6/6 is not seen as exceptional or genius]. You should actually prefer odd numbers of points because they make it easy to give somebody a neutral score on a symmetric scale. The study authors used 6 points because one of their two experiments was based on pre-existing data, and that is how many points the existing data had.