I'd actually be interested to see if there is a prevalence to use one gender over another for imaginary people. Not sure what it would mean, but it'd be interesting. Even more interesting if there was comparable data for people in other industries.