where Y is the movie revenue (in millions of dollars) and X is
movie budget (also in millions of dollars), using all
observations to estimate the model (the restricted model).
Then estimate the model separately for movies with a female
300
lead and those with a male lead (two unrestricted models).
Use these regression results to calculate the Chow F-test.
Figure 9-12 contains the STATA output needed for
performing a Chow test — the restricted regression with all
movies and the two unrestricted regressions (one using the
sample of movies with female leads and one using the sample
of movies with male leads).