I doubt that scientifically reputable research into this would is possible. How would you design the study? Start a bunch of successful large corporations then

I doubt that scientifically reputable research into this would is possible. How would you design the study? Start a bunch of successful large corporations then randomly staff their executive positions with different mixtures of men and women, and study the effects on society at large for 30 years?