1. Consider the following table, where the unit of observation is an individual, and the dependent variable in each regression is annual income in dollars. Each column represents a separate regression. The standard errors of the coefficients are reported in parentheses. Female is a dummy variable which equals 1 if an individual is female; and equals 0 otherwise. [15 marks]

Dependent Variable: Income in Dollars	Model 1	Model 2	Model 3	Model 4
Years of Schooling	4362.88***	4364.63***	5243.88***	5188.76***
	(1214.52)	(1242.74)	(1439.94)	(1625.75)
Female			-2324.86***	-2924.86***
			(861.57)	(1061.27)
Female X Years of Schooling				1126.83***
				(323.42)
Province Fixed Effects	No	Yes	No	No
Adjusted R-squared	0.256	0.253	0.297	0.311
Observations	106,222	106,222	106,222	106,222
* p<0.10, ** p<0.05, *** p<0.01

1. Do province fixed effects explain any variation in the dependent variable? [1 mark]
2. Does including province fixed effects mitigate any omitted variable bias concerns? [1 mark]
3. Interpret the two coefficients in model 3. [3 marks]
4. The variable Female is negatively correlated with the dependent variable, and positively correlated with the variable Years of Schooling. In light of this, explain the change in the coefficient on Years of Schooling from Model 1 to Model 3. [1 marks]
5. Interpret the coefficient on the variable Female X Years of Schooling in model 4. [3 marks]
6. What is the total effect of one additional year of schooling on annual income for women, based on the results shown in the table? [3 marks]
7. Consider another dummy variable Male, which equals 1 if an individual is male and equals 0 otherwise. If you accidentally added this variable to the regression in model 3, which OLS assumption will be violated? [1 mark]
8. Describe a visual test that you would conduct to examine whether errors are heteroskedastic in model 1. When will you conclude that errors are homoskedastic? [2 marks]