Regression of gender on income

Genders are Male & Female
Entrepreneur is given by dummy entrepreneur (1)
Income is based in 16 groups (0-10k; 10k-20k; 20k-30k etcetera)?

How can I approach this with an OLS method and for what should I test (correlation, homoskewdasticy etcetera)?

I know that there are other factors and I will test for them as well.
Thanks for the help!


Less is more. Stay pure. Stay poor.
ideally your outcome is continuous. Why aren't you using continuous salary, you are losing information by categorizing them.

Also, big debate in areas where age or say race are examined. In that these are just proxie s to the real latent variable at play. Meaning you cannot change these variables.
Last edited:


Active Member
If your dependent variable is binary, OLS regression is not likely to be valid. Consider using logistic regression instead.