I was making a counterpoint to the parent post. Note the links there, nothing much to do with tech. And you're completely wrong about sample sizes:

"Comparisons of oral non–gender-blind tests with written gender-blind tests for about 100,000 individuals "

"Over 2,100 public servants"

"validation studies were conducted on 873 tenure-track faculty"