STAT 601: Applied Statistics I
FOR THE COURSE
Class meets: MW 4:00-5:15 pm in:
Instructor: Ranjan Maitra
Office: Mathematics and Psychology, (MP) 436
Office Hours: By appointment.
Systems Administrator: Susan Burgee
Office: Mathematics and Psychology, (MP) 413
Computer accounts on math.umbc.edu may be obtained by filling
in the form posted outside Ms. Burgee's door.
STAT 601 is an introductory graduate level course on Applied Statistics
and Data Analysis. The course is intended to be an introduction to the
real world of statistics, focusing on perhaps the most widely used tool
in statistical science -- linear regression. Ranging in applications from
econometrics to medical imaging, linear regression provides a language
and a body of techniques for specifying and measuring the quantitative
relationships between inputs and outputs in a system. We will look at real
data, try various models for the data, assess the validity of assumptions,
and try to arrive at conclusions. Computer programs will do most
of the calculations leaving us to concentrate on the analytical aspects.
The course will therefore primarily be structured around the theory and
applications of the linear model. We will start with introductory material
and then cover Chapters 2 through 8, and then, time permitting, go on into
more advanced topics such as Generalized Linear Models, etc. While we will
follow the topics in the book for the most part, I will also supplement
examples from other sources.
A statistician's job is not complete when the last decimal point
is calculated and the last graph is drawn, however. Statistical work is
almost always done in collaboration with a researcher in another field
-- perhaps economics, engineering, medicine, pyschology, etc. -- and communicating
the data analysis effectively to others is as critical as doing the data
analysis well to begin with. When you communicate well, your work influences
the research in the other field and you get credit. When you do not, your
work goes unnoticed and your hard work goes to waste. A substantial
part of the course is therefore going to be on data analysis and report
writing, especially for clients whose knowledge of statistical tools and
methodology borders on the rudimentary.
The package that will be used in the course will be Splus[available
on sgi1 and suns in math, and on gl and research in umbc], though you are
welcome to use other packages such as SAS [also available on sgi1 in math
and on gl in umbc] or SPSS in your assignments. This is not a computer
course, so we will not actually go into how to write Splus code, but you
are always welcome to contact me should you have any questions. Also, I
will probably arrange for a guided tour of Splus in the Stat/Math lab after
the class strength has stabilized.
There will be two exams during the course of the semester. I will
set dates for the exams in consultation with all of you. The score on the
examinations will contribute to 35% of your final grade.
Homeworks will be handed out almost every class, or every other
class. The homeworks will cover both theoretical and applied problems.
It is assumed that graduate students will arrive at homework solutions
fairly independently, though discussions and joint exploration of the computer
system is always welcome. [It is not necessary that computer assignments
be handed in Splus: you are free to use whatever resource you are most
comfortable with and whatever does the job, as long as that resource is
not your friend's assignment!] The homeworks together contribute 35% of
your final grade for the course.
There will be two projects assigned during the semester. The purpose
of these projects is to enable you to demonstrate that you are able (a)
to frame the problem in a scientific context, (b) to apply the statistical
methodology that you have learnt in the class, and (c) most importantly,
to communicate effectively with clients who have only an elementary knowledge
of statistics. I will personally guide you through the first project
by means of class discussions. On the second project however, you will
be expected to work more independently, even though there will be class
discussions occasionally. A 10-15 page report on the first project, to
be handed out in the second week of classes will be due on or before October
19, 1998. The report on the second project to be handed out by the end
of October will be due the last day of classes. The two projects together
contribute 30% of your final grade. Grading guidelines for these reports
will be detailed separately.
Here is a sample
For faster communication, I would like to maintain a class e-mail
list. Please therefore, send me an e-mail at my e-mail address at email@example.com.
Be sure to tell me who you are, your major,
background, etc., what are the things that interest you [outside
statistics], what you expect to get out of this course. This will
help me to modify the course structure if necessary and also choose examples
that some of you may find more interesting. Also let me know how often
you check your e-mail. (Though it may not always be possible, I urge that
you check it at least once a day.) Of course, feel free to criticise and
make suggestions about the class at any time.
The course homepage will be located on the WorldWide Web at:
Assignment 1: The
I will try and keep this homepage as upto date as possible. However,
you are still responsible for any announcements made in class.
Assignment 2: hooker
Assignment 3: salary
Assignment 4: haus
Assignment 5: lmdiag
Project 1: problem description and ortho dataset
Project 2: problem description and sampdly dataset
Final: problem description; universities data
and universities documentation ; Academe data and Academe documentation