### Teaching

Instructor: Russell Davidson. TA: Vinh Nguyen
• Marks: you can find the marks yourself. Just follow the steps below.
• Install MATLAB, which is freely available for McGill students; if you have MATLAB already, ignore this step
• UPDATE 2016-10-12: initial version with marks for Assignment 1
• UPDATE 2016-10-28: redesigned version with marks for Midterm added
• UPDATE 2016-12-25: with marks for Assignments 2 and 3 and Final added
• UPDATE 2017-01-04: with Q6 removed from raw score and with normalising denominator fixed at 70
• In MATLAB, set MATLAB directory to the two downloaded files
• Type help grades in MATLAB to learn how to use the files;
• Example: type grades(123456789) to see marks for student with ID 123456789
• Remark: before an exam has taken place, everyone gets 0 for that exam; that will of course be fixed once the corresponding exam has happened and has been graded
• Assignment files
• Midterm
• Grading progress: complete; see grading notes
• Descriptive statistics: max = 97.22%, mean = 42.08%, std = 27.82%
• Final
• Grading progress: complete!
• How the final is marked: each student receives a raw mark based on the student's performance on the exam. This raw mark is a nonnegative number between 0 and 96. The student's effective final score is then the ratio between the raw mark (student-specific) and a normalising denominator (common to all students).
• Normalising denominator: 70
• Descriptive statistics for the normalised scores (among students who actually wrote the exam): max possible = 100%, max actual = 100%, avg = 62.39%, std = 20.56%
• Let F be the normalized final score
• Let M be the midterm score
• Let A be the average among 3 assignments
• The course grade G is computed as: max{F, (F+M)/2, (F+(M+A)/2)/2}
• Note: F, M, A, and G are all measured in percents
• Normalized final v.s. midterm: (this only includes students who actually wrote the final) (so this is a plot between F v.s. M, NOT G v.s. M)

• Notes

McGill, Winter 2016, Econ 665 (Quantitative Methods)
Instructor: Vinh Nguyen

McGill, Fall 2015, Econ 468 (Honours Econometrics I)
Instructor: Russell Davidson. TA: Vinh Nguyen
• Syllabus
• MATLab files (built on MATLab 2015b x32):
• Assignment 1 files (link removed as course is over!)
• Assignment 2 files (link removed as course is over!)
• Assignment 3 files (link removed as course is over!)
• Assignment 4 files (link removed as course is over!)
• Assignment 5 files (link removed as course is over!)
• Notes from OH:
• 2015-11-12
• 2015-11-19 (an intro to tests of hypotheses)
• 2015-11-26 (book exercises 4.2 and 6.7)
• 2015-12-03 (notes on projection matrices and book exercises 2.16 and 2.19)
• 2015-12-10:
• Part 1: solutions and hints for Exercises 4.4, 4.6, 4.8, 4.9 (removed as course is over!)
• Part 2: (unedited!) notes from the tutorial (a data question on seasonality effects and a question about bootstrap tests in the context of IV regressions)
• Other notes:
• Nonparametric bootstrap through an example: rescaling residuals and other issues
• Due to time constraints, a planned review on GNR is cancelled. Sorry!
• The final is out of 154. Your relative score for the final is computed with this formula: min{(raw score / 100),1}. In short, you are scaled by 100 (instead of 154) and if your relative score exceeds 1, it is rounded down to 1. The effective denominator 100 is chosen by Professor Davidson and the raw scores are supplied by me.
• Histogram for relative scores on the final exam:

• Histogram for the overall performances in the course (among those who have written the final exam):

• Regressing the relative score on the final exam and the final grade on a constant, homework average, and midterm score yield:

 Final exam Final grade Coefficient p-val Coefficient p-val Constant 0.2025 0.0336 0.1976 0.0034 Homework 0.2421 0.0438 0.2371 0.0052 Midterm 0.4423 0.0052 0.5339 0.0000