RPP = β0 + β1 DIST + β2 RM + ε
where
RENT = Total monthly rent in dollars
NO = Number of persons in apartment
RM = Number of rooms
SEX = 1 if female; 0 if male
DIST = Distance from center of campus, blocks
RPP = RENT/NO = Rent per person
(1.1) Do you find any evidence of sexual discrimination in this set of rental data based on the estimated regression equations? Answer this question using two approaches: sample separation approach (Chow test) and dummy variable approach.
(1.2) One of the advantages of sample separation approach is that the difference of regression variances for male and female sub-samples can be estimated and tested explicitly. Setup and perform the appropriate test procedure for the change in regression variances in between the estimated male and female equations.
This homework is based on the classical study (paper):
Solow, R. "Technological Change and the Aggregate Production Function." Review of Economics and Statistics, 39, 1957, pp.312-320.
The Solow original data (several variables are omitted) is avialable here (see also, Greene's Table F6.4):
Four linear regression models were considered:
Formulate and perform appropriate tests and compare the above four non-nested models. Which model is a better one? Explain.