Heckman 2 step procedure - how to formulate probit model for real estate?

#1
Dear all,

I'm doing research on the attributes that determine listing prices of houses. Price is updated to the current date and independent variables are structural (e.g. surface, building year), related to the neighborhood of houses (e.g. density of population in neighborhood, distance to supermarket, distance to school, % single family dwellings) and environmental (e.g. average gas usage and average electricity usage in neighborhood).

My professor told me to use the Heckman 2 step procedure to account for sample selection bias. I've been reading a lot on it. However, I cannot figure out how to formulate the second (probit) equation. First equation is as described above: Price_i = alpha + beta(Si) + gamma(Ni) + delta(Ei) + error. How to formulate the second equation? Do you got any ideas?
I thought of: listing (yes/no), but then, I don't have any independent variables to test for this probit equation. Maybe I do not perfectly grasp the intuition of Heckman 2 step procedure yet.
Calculation in R would not be a problem.

Could you please give me any insights?

Thanks in advance!

RealEstate