Why is the fact that the coefficients on SAT scores, and parent's income disappear once you control for the group dummies?
a) It suggests that the group dummies are probably a good proxy for unobserved background variables (since they are a good proxy for the observed variables).
b) It suggests that there may be other omitted variables.
c) It suggests that our measures of SAT score and parent's income are very imprecise.
d) It tells us that SAT score and parental income have no impact on earnings.
