LINEAR MODELS 2018-2019

Master of Statistics
University of Hasselt

Student manual

1 Introduction
This course introduces the student to simple and multiple linear regression mod-
els, including one-factor and two-factor ANOVA models. The student will learn
when it is appropriate to apply these techniques, how to implement them, how
to interpret and communicate the results. The student will also learn how to use
diagnostic tools to evaluate the goodness of fit of these models. He/she will also
be able, at the end of the course, to apply remedial measures when necessary,
in order to get reliable and meaningful results.
The basic ingredients are introduced in the courses Concepts of Probability
and Statistics. The implementation of the methods is shown using the software
R and SAS. Every student will study the course contents individually, but col-
laboration with other students is very much encouraged. The lectures will be
given in a mixed video and live format and non-mandatory exercises will be
available to the students to practice in preparation of the final exam.
The evaluation is permanent: the student will be evaluated both via quizzes
and homeworks, during the lectures period, and via a final written exam.
Advised prerequisite of this course is basic knowledge of fundamental sta-
tistical concepts (Concepts of Probability and Statistics course). Moreover, the
knowledge of basic mathematical concepts is crucial. The student is expected
to have an appropriate background knowledge of mathematical concepts, as in-
dicated by the competencies matrix of the Maths&Stats module. If the student
lacks any needed knowledge, he/she is expected to catch up on it before the
start of the course, via the Maths&Stats platform.
All course material (handouts, video lectures and exercises) will be available
on Blackboard in the “Course Material” and “Assignments” folders. In Section
2, the structure of the course is outlined. More details about the course material
are available in Section 3. Section 4 indicates how the student will be evaluated
at the end of the course.

1
2 Structure of the Course
Most of the concepts discussed will refer to chapters in the book Applied Linear
Statistical Models, 5th Edition, Kutner et al. (2005) and Applied Multivariate
Statistical Analysis, 6th Edition D. W. Wichern and R. Johnson (2007). In this
section, a summary of the course structure is presented.

2.1 Lectures
The lectures will be provided in both video and live format. Multiple videos per
lecture will be available, each of which will summarize a specific section within
the lecture. A description of the content of each videos is available on Black-
board in the “Course Information” folder (‘LINEARMODELS videos summary.pdf’).
The videos will be made available during the semester, following the same
time schedule as the live lectures. The video lectures can be used for self-study
and review.

2.2 Exercises
Exercises will be made available during the semester. They will be related to
sets of lectures and they will be available right after the videos for these lectures
are accessible. The student is expected to solve the exercises and check their
results with the solutions, which will be provided in the form of R/SAS code.
By default, the exercises will not be discussed in class. However, the student
can ask questions related to them during any of the lectures.
An example of final exam will be provided on Blackboard. Two lectures
during the semester will be dedicated to the discussion of this example. The
student is expected to answers on his/her own the questions in the exam example
before these lectures. During these lectures, students will be randomly selected
to discuss/share their solutions.

2.3 Quizzes
There will be two multiple choice quizzes on scheduled dates during the semester,
which are reported in the key dates document. These intermediate evaluation
tools will serve to test the student on the concepts taught in the lectures.
These quizzes will be evaluated and the score will count up to 1.5 points for
the final mark. The quizzes are individual, thus the student is expected to solve
them on his/her own and not in collaboration with other students. Feedback
on the quizzed will be provided immediately after each quiz.
The student will take the quizzes on campus during the first semester. If the
student will not be present for the quizzes, he/she will loose the related points
for the final score.
The content of the quizzes will be:

1. Quiz 1: Parts 3 and 6

2
2.4 Homework
There will be two homework on scheduled dates/deadlines during the semester,
which are reported in the key dates document. Dataset for the homework is
taken from the book Kutner, et al.
To facilitate collaboration with your colleagues it is highly recommended
that, for the two homework, you work in groups (four or five is good). In addition
to creating a good environment for helping each other, this also develops a team
working spirit which is crucial for a statistician.
Each homework has a specified deadline and it is expected they will be
respected. Homework assignments can be submitted before or at the designated
time. Prepare your report/solutions in latex and submit only a printed copy
of the pdf file which can be placed, unless otherwise specified, in the lecturer’s
mailbox. Use of statistical software is allowed.
The two homework will count up to 1.5 points for the final mark. Feedback
on the homework will be provided during specific discussion lectures.
The content of the homework will be:
1. Homework 1: Parts 1, 2 and 4
2. Homework 2: Parts 5, 7 and 9

3 Course material
The following course material is available on Blackboard:
• The slides/handouts and the videos for the each lesson are available on
Blackboard in the “Course Material” folder, at the scheduled dates.
• The R and SAS code to deal with all taught concepts is provided in the
slides/handouts.
• All the exercises and the R/SAS code with the solutions are available on
Blackboard in the “Assignments” folder, at the scheduled dates.

4 Evaluation
The course will be graded on the basis of:
A. Quizzes: Each of quizzes 1 and 2 will be given a mark from 0 to 0.75. For
the quizzes, the mark will be obtained by counting the correct answers in
proportion to the total number of questions. Wrong and missing answers
will be counted as 0 scores.
B. Homework: Each of the two homework will be given a mark from 0 to 0.75
points. Since this will be a group work, all students within a group gets the
same mark. Copying from or collaborating with another group is strictly
prohibited. Any groups caught doing so will be given a mark of 0.
C. Written exam: the final written exam, consisting of two components (ANOVA
and Regression), will count up to 17 points for the final mark. In this exam,
real data problems will be presented to the student, who will have to answer
questions about both theoretical and computational aspects.

3
 Practically, the division of the total mark is:

quiz 1 [0, 0.75]

total [0, 20]

This division will be applied both for the first and second chance evaluation.
No second chance quizzes or homework will be offered.

Important Note:

For the student to pass the course, a total mark of at least 10 is needed wherein
for the final written exam, the student should score a minimum of
35% (or 2.975 out of 8.5) for each of the components. In case a score
below 35% is obtained for at least one of the two components, the total score
of the student will be the minimum between:
• 9 or
• the sum of all evaluation forms of the course.