You are on page 1of 4

# STAT3008 Course Outline

## Course title: STAT3008 Regression in Practice 1st semester, 2017-2018

Instructor: Professor CHEUNG Siu Hung [Lady Shaw Building 117, phone (off.): 39437928]

Time and Venue (Lecture): Tuesday 1:30-2:15pm [Lady Shaw Building LT1]
Thursday 1:30-3:15pm [Lady Shaw Building LT1]

Description:

The course introduces regression methodologies. Topics include the use of correlation coefficient as a
measure of relationship, the use of simple linear regression, multiple regression and logistic regression in
projection and forecasting, as well as the use of model building techniques to incorporate qualitative
variables in prediction.

2. Learning outcomes

## Upon completion of the course, students will

a) learn the essential skills in estimation and statistical inference using linear regression models,
b) employ appropriate techniques in regression model building,
c) understand the procedures of applying regression analysis to real datasets in practical problems,
d) become familiar with the SPSS statistical software (the regression module),
e) be able to use the regression model for prediction and forecasting.

3. Pre-requisites

## a) Elementary Statistics and basic distribution theories

b) Basic knowledge in calculus and linear algebra

4. Course Content

## Topic Contents/ concepts

Simple Linear Regression Simple linear regression and the method of least squares,
estimation and inference of regression parameters, interpretation
of the regression equation, prediction and residual plots
Multiple Linear Regression Introduction of the multiple regression model, least squares
estimation, maximum likelihood estimation and inference of
regression parameters, regression models for quantitative and
qualitative predictors, prediction and applications
Model Building Model building and variable selection, stepwise regression and
all-possible-regressions selection procedure
Diagnostics and Regression Diagnostics, residual analysis, multicollinearity, remedial
Pitfalls measures
Weighted least squares regression Estimation, inference and applications
Logistic regression

Page 1 of 4
STAT3008 Course Outline

4. Learning activities

## Lecture [(in/out) - 3hrs/3hrs]

Computer Lab

5. Assessment Scheme

## Type Description Weight

Assignment Around 8 sets of assignments 15%
Test 1 1 A-4 size formula sheet (written on both sides) allowed 20%
Test 2 1 A-4 size formula sheet (written on both sides) allowed 20%
Final Examination 2 A-4 size formula sheets (written on both sides) allowed 45%

## 6. Feedback for evaluation

Students comments and feedback are valuable for improving the course, and students are welcome to
provide comments and feedback via various channels. Mid-term course evaluation and course evaluation
will be conducted for the course. For comments that require immediate action, please send to Professor
SH CHEUNG directly via e-mail (shcheung@cuhk.edu.hk) or to the teaching assistants.

## Professor/ Lecturer/ Instructor

Name: Professor CHEUNG Siu Hung
Office Location: LSB 117
Telephone: 39437928
Email: shcheung@cuhk.edu.hk
Personal Website: http://www.sta.cuhk.edu.hk/shcheung/

## Teaching Assistants: CHU Dingnan [Lady Shaw Building 128, 1155088229@link.cuhk.edu.hk]

GAO Linhao [Lady Shaw Building G28, 1155090017@link.cuhk.edu.hk]
TANG Wenlu [Lady Shaw Building G30, 1155082898@link.cuhk.edu.hk]

Page 2 of 4
STAT3008 Course Outline

8. Course schedule

## Date (2015) Topic

Sept 5 Brief overview of the course
Arrangement of tutorials
Sept 7, 12 Review: Normal distribution, sampling distributions of the sample mean
and sample variance, t-distribution, maximum likelihood estimation
Pearson product moment coefficient of correlation (measure of linear
relationship between two quantitative variables)
Sept 14 Introduction to simple linear regression
Model assumptions
Least squares estimation
Sept 19 Maximum likelihood estimates and their distributions
Sept 21, 26 Statistical inference in simple linear regression (confidence intervals and
hypotheses testing)
Interpretation of the regression equation
Prediction
Regression through the origin
Residual plot
Sept 28, Oct 3 Introduction to multiple linear regression
Multiple regression in matrix notation
The model
Least squares estimation
Maximum likelihood estimation
Oct 5 Holiday: Mid-Autumn Festival
Oct 10 Analysis of variance, the F-test and the coefficient of determination
Oct 12 Test 1 (on simple linear regression)
Oct 17 Analysis of variance, the F-test and the coefficient of determination (cont)
Oct 19, 24 Estimation and inference of parameter in multiple regression
Partial F-test
Interpretation of parameters in regression models
Oct 26, 31 Quantitative and qualitative predictors
Prediction and applications
Nov 2, 7 Model building
Nov 9 Test 2 (on multiple regression before model building)
Nov 14 Variable selection
Nov 16 Congregation, NO Class
Nov 21 Diagnostics and regression pitfalls
Remedial measures
Nov 23, 28 Logistic regression
Nov 30 Revision

Remark: The exact schedule for assignments will be announced in class and in the course website.

Page 3 of 4
STAT3008 Course Outline

## 9. A facility for posting course announcements

You may use the University e-learning platform to access course information and course materials. The
website is managed by the University. Should you have enquiry on the system, please contact Mr. Julian
Wong (julianwong@cuhk.edu.hk) by e-mail. Should you have comments on the course content, please
contact Professor SH CHEUNG directly (shcheung@cuhk.edu.hk) or contact the teaching assistant.

## (1) Lecture notes

For students easy reference, lecture notes have been prepared in the format of power point files
or pdf files. Students can download the files from the University e-learning platform: Blackboard
Learn.

## (2) Regression Using SPSS

Videos that demonstrate the use of SPSS can be downloaded from university Blackboard Learn.
Students will be able to view these files with Internet Explorer.

(3) Assignments

All assignments and their due dates will be posted on the University e-learning platform:
Blackboard Learn.

## Mendenhall W and Sincich T. A second Course in Statistics: Regression Analysis,

Pearson/Prentice Hall

## Honesty for Academic Works: A Guide for Students and Teachers

Attention is drawn to University policy and regulations on honesty in academic work, and to
the disciplinary guidelines and procedures applicable to breaches of such policy and
regulations. Details may be found at http://www.cuhk.edu.hk/policy/academichonesty/ .

Page 4 of 4