Cross Sectional Analysis in More Than One Dimension

The following personal experience represented a milestone in my thinking about regression, and in fact, about socio-economic data in general. The data were from the sex-discrimination lawsuit,¹ in which I was a statistical expert-witness for the plaintiff. The group of female employees claimed to be paid less than male librarians at entry level, to have received smaller pay raises and fewer promotions, and yet to have the same qualifications and identical work requirements as their male counterparts. The plaintiff’s lawyer approached me to explore and statistically establish the facts of their claim. The District Court ordered the unrestricted access to the employment records of the professional men and women in that large agency of the Federal Government.