Women are one of the under represented demographics in amateur radio. This is evident in meetings, on the air, and at tests. Finding demographics such as number of males and females or their ages is hard because the FCC does not ask them in the application for a license.
One way to approximate them is to make a best guess based on the name. The US Census Bureau has a list of male and female names at https://www2.census.gov/topics/genealogy/1990surnames/. Each list contains the name, percentage with the name, cumulative total, and rank.
One thing that we notice is that there are some names that appear on both lists. Some are safe to assume a gender: James(3.318% of males and .01% of females) and Mary(2.629% of females and .009% of males). Others are harder to determine: Kris (.011% of males and .011% of females) and Sydney (.007% of males and .008% of females)
We can make a guess of the gender if the name has a much higher percentage of one gender than the other. Choosing 5x or 10x gives pretty good guesses for the gender.
Assigning a gender to all the names in the FCC database gives the following results:
We find that 9% of the names either aren't on either list or are less than our threshold to guess. We can throw out the unknown genders assuming that the proportions are about the same as the known genders. This gives the folowing results:
This results in 86.14% of hams being male and 13.86% of hams being female. These numbers are close to similar studies done such as one done by Ken Harker WM5R in 2005.