Dear students,

Below is the information about the final exam. Try to read it carefully to avoid any mistakes:

1/ Exam dates and time:

·         Exam 1: 20 December (9:10 – 11:10, register at LINK 1 ends by 23:59 on 18 December);

·         Exam 2: 10 January (9:10 – 11:10, register at LINK 2 ends by 23:59 on 8 January);

·         Exam 3: 17 January (9:10 – 11:10, register at LINK 3 ends by 23:59 on 15 January);

·         Exam 4: 24 January (9:10 – 11:10, register at LINK 4 ends by 23:59 on 22 January);

2/ Exam structure:

·         Final exam will account for 40% of the total grade;

·         Final exam will last for 2 hours;

·         Final exam will be structured with MCQ (30 pts = 10 * 3pts), Theoretical questions
(choose 2 out of 4 questions, 30 pts = 15 * 2pts), and Practice Exercises (40 pts);

·         Final exam covers all contents from Lecture 1 to the final lecture of week 13 December;

·         You are allowed to bring ONE A4 (2-sided) cheat sheet with HANDWRITTEN notes to use in
the final exam;

·         You are suggested to bring a calculator to use in the final exam;

3/ Theoretical topics to review:

·         OLS classical assumptions (name, detailed explanation);

·         4 criteria to choose a variable in the model;

·         Dummy variables (types, model, interpretation);

·         Omitted variable and irrelevant variables (issues explanation, consequences, remedies);

·         Multicollinearity and heteroskedasticity (issues explanation, consequences, tests to
detect, remedies);

·         Autocorrelation (issues explanation, consequences, tests to detect, remedies);

·         Endogeneity (issues explanation, consequences, causes, remedies);

·         Binary dependent variable model (up to slides 16)

4/ Practice exercises to review:

·         Write down the model, numeric model with regression output provided;

·         Interpret coefficients in multiple functional forms;

·         Interpret coefficients of dummy;

·         Interpret coefficient of determination (R-squared);

·         Perform different types of tests (all tests can be asked, no exception) and interpret the
result of accept or reject H0;

·         Discussion about the impact of omitted variables (conduct omitted variable bias OVB);

·         Discussion about the impact of heteroskedasticity with appropriate tests might be asked;

·         Discussion about the impact of autocorrelation with appropriate tests might be asked;

·         Some calculation with given values of explanatory variables.

Should you have any questions, feel free to send me an email.