Math 3200 - Spring 2013

Statistics and Data Analysis

Section Information

Section Time Location Instructor Office Hours (Cupples I, Room 17)
1 MWF 1:00 PM - 2:00 PM Brown 118 Renato Feres MWF 2:00 PM - 3:30 PM

Please include [Math3200] in the subject line of any email message that pertains to this course. This will help avoid that I accidentally delete your still unread message. My e-mail address is feres@math.wustl.edu. If your email message is concerned with a suggested exercise, cc it to the teaching assistants (see below) in addition to sending it to me; this way, the first person available will be able to answer your question.

Teaching Assistants

The Math 3200 TAs this semester are:

Re-I Chin
Gloria Han
Benjamin Misch
Ningyuyuan Wang

In addition to helping with grading homework assignments, our undergraduate teaching assistants will be avaliable for a few hours of the week at the help desk at Cornerstone.

Text

Statistics and Data Analysis by Ajit C. Tamhane and Dorothy D. Dunlop
Printice Hal, 2000

The textbook webpage is http://www.prenhall.com/tamhane.
You'll find there a link to an errata file (or here) and a link to data file downloads (also here).

Topics covered

This is an introduction to probability theory and statistical analysis. Topics to be covered include an overview of the basic ideas of probability, discrete and continuous random variables, hypothesis testing and inference, summarizing and exploring data, linear and multilinear regression, single factor anaylsis of variance, among others.

Statistics software

This course will use statistics software in an essential way. Software for statistics computing in Math 3200 varies, but in the past the language of choice has been SAS. For reasons that will be more fully explained in class, this semester we are going to rely exclusively on R. The R language is widely used for professional statistical work, especially scientific and academic research. It is very powerful and flexible, and is entirely free. Tutorials and specific information at all levels of usage, from the installation process to sophisticated applications, are easily available online.

In the R computing main page you'll find instructions for downloading and installing R and general documentation. In particular, the manual An Introduction to R is a useful source of information.

Although the plain R program is nice enough in my opinion, there are free programs that provide more friendly user interface environments in which to run it. For example, I find that RStudio looks especially nice.

I do not assume that you know R. The information about it provided in homework assignments and occasionally in class should be enough for the needs of the course.

Grading Information

There will be three evening exams during the semester, E1, E2, E3, and a final, E4.
These are multiple choice exams. I'll tell you more in class about their general format. Times and locations are as follows:

Exam Date Location Time Solutions
E1 Wed, 02/06 Check here on exam day. 7:00-9:00PM Exam I - Solutions
E2 Wed, 03/06 Check here on exam day. 7:00-9:00PM Exam II - Solutions
E3 Wed, 04/10 Check here on exam day. 7:00-9:00PM Exam III - Solutions
E4 Thur, 05/02 Check here on exam day. 3:30PM-5:30PM Exam IV - Solutions


Your final grade will be based on the three exams, E1, E2, E3, the final exam E4, and the homework score, W. I expect there to be a total of 10 assignments; the worst HW score will be droped in calculating W. These five scores are scaled so that each is worth a maximum of 100 points. They are then combined according to the following formula:

S = 0.75*(E1 + E2 + E3 + 2*E4 - min(E1, E2, E3, E4))/4 + 0.25*W



The term in parenthesis, T=E1 + E2 + E3 + 2*E4 - min(E1, E2, E3, E4), says that the final exam score, E4, replaces the worst score among E1, E2, and E3, if E4 is not the least value (so that E4 effectively counts twice). If E4 is the least score, then all four exams are counted with equal weight. Then the total exam score T/4 and homework W are added up with weights 75% and 25%, respectively, to obtain the total score S.

The value of S will be translated into a letter grade of A, B, C, D, F (with plus and minus shadings) in a way that is not stricter than the following table (if the grade distribution of the whole class is significantly lower than usual, then "curving" may be considered, but it is unlikely to be needed):

Numerical Range Letter Grade
[85, 100] A
[70, 85) B
[60, 70) C
[50, 60) D
[0, 50) F


For example, if the evening exam scores are 89, 60, and 81, the final exam score is 94, and the assignments total is 97, then 94 replaces the 60, and

S=0.75*(89 + 60 + 81 + 2*94 - min(86, 60, 78, 94))/4 +0.25*97 = 91.40.

This gives a grade somewhere in the A range.

In assigning pluses and minuses I will observe the following policy: A+ is reserved for the very top scores, and is given sparingly. I expect that only a very small number of students will received an A+. The ranges of letters B and C will be divided roughly evenly into +, -, and unsigned grade, but the exact ranges will take into account the full set of class scores so as to minimize borderline cases. Therefore, these ranges will not be determined until after the full set of class scores are known at the very end of the course.

Exams


Exam questions are going to be drawn from the list of suggested problems (third column of the course plan below) after some modifications, so make sure that you have solved those suggested problems prior to a test. These problems may not appear in a given test with exactly the same wording or numerical values as in the book since it is always necessary to restate them in an exam appropriate format, taking into account time limitation and other factors.

Exams typically will consist of 10 multiple choice questions. (Amended: Exams will be based on no more than 10 problems of the type given below in the suggested list. However, in order to allow for partial credit, I may break up the longer problems into a few smaller ones, so that the exact number of exam questions may exceed 10.)

If you know the material well, there will be enough time to work out test problems carefully and without rush. Guessing won't be a good strategy. Although the calculations leading to the result of a test problem won't be graded, it is always a good idea to write them carefully, in a clear and legible way. This not only reduces the chances of making simple mistakes, but will allow you to contest the score if you think there might have been a grading error.

Exam question can be done by hand with the help of a basic scientific calculator. (In particular, you don't need a statistics calculator for this course.) The allowed calculators will be the same that can also be used in Calculus exams. (They shouldn't be graphing calculators.)

Here are a few general items to keep in mind:

You should always bring your Washington University Photo ID to exams. Proctors will check student's IDs.

All exams are close-notes-and-close-book. However, a 4" x 6" index card may be used as a "cheat sheet." You may write on both sides and include any information you deem useful.

Calculators are recommended, but will mostly be used for simple arithmetic. All the distribution tables you may need (which appear in the appendices of our textbook) will be provided on the exam, so your calculator need not have any specialized statistics capabilities.

Just before each exam, look up your exam room assignment on the web at this link. The room will likely not be your regular classroom, and it may be a different one for each exam. You will be allowed to enter the exam room a few minutes before the starting time to locate your seat and exam booklet.


The Mathematics Department has the following policy in place for the large sections regarding date of the final exam and excused absences:

The exam dates including the final were set by the College office before you registered for this class. You are therefore expected to take the exams at their scheduled times. If you are away because of a university sporting event or field trip, then you may arrange for your coach or professor to administer the exam. Excused absences may be granted in the case of illness or bereavement. All excused absences must be granted by Blake Thornton.

The final exam date cannot be changed for reasons of traveling convenience.

Homework

The plan is to have a total of 10 HW assignments, mainly involving computer work using R. Of these 10, the 9 best scores will count towards the total grade score W. (See the formula above.)

You are allowed and encouraged to collaborate on your assignments. This is especially useful when dealing with issues related to the software, such as figuring out what R commands to use to compute this or that statistical function or plotting a desired type of graph. If you worked in a group on a given problem set and feel confident about the solutions agreed on by the group, it will be fine to submit a single copy signed by the group members. This way, the grading work by the undergraduate TAs is reduced (thus freeing them more time that they can use for answering questions and mentoring). The group does not have to be the same for all assignments.

Assignments are to be turned in on the due date (Fridays) at the beginning of class. Late assignments won't be accepted.

Course plan

Visit this course plan on the on-line syllabus regularly to find the latest entry on suggested problems and other useful course information.

Your main course activity, in addition to homework sets, will be solving the below suggested problems. Although these problems (all from the textbook) are not collected, they will be my source of test problems. Make sure that you know how to do them with confidence in preparation for a test. In fact, the best way to get ready for a midterm test or final exam is to carefully solve all those problems on the topics covered in a given exam, reviewing the theory as needed.

Please keep in mind that the plan given below is somewhat tentative. Sections and problems may occasionally be changed, and more links will become active as the course progresses. Material covered in each test will be detailed in class.

Week Chapter Suggested Problems
Jan 14 - Jan 18
Homework 0
(not for grade; won't be collected) 		2 2.1: 6, 9, 10, 11, 12, 14
2.2: 16, 17, 18, 20, 22, 27
Jan 21 - Jan 25
Homework 1: 1/25
Solutions 		2 2.3: 28, 29, 30, 32, 33, 34
2.4: 35, 38, 39, 40, 41, 42
Jan 28 - Feb 01
Homework 2: 2/01
Solutions 		2 2.5: 48, 49, 50, 52, 53, 54
2.7: 59, 60, 61, 62, 64, 70
Feb 04 - Feb 08
EXAM: Wed, 2/06
Solutions 		2 2.8: 71, 72, 73, 74, 75, 76
2.9: 78, 79, 80, 81, 82, 83
Feb 11 - Feb 15
Homework 3: 2/15
Solutions 		4 4.1: 2, 3, 4
4.2: 5, 6, 8
4.3: 9, 10, 11, 12, 14
Feb 18 - Feb 22
Homework 4: 2/22
Data
Solutions 		4, 5 4.4: 29, 30, 31, 33, 34, 38
5.1: 1, 4, 6, 7, 8
5.2: 16, 18, 19, 20, 22, 23
Feb 25 - Mar 01
Homework 5: 3/01
Solutions 		5, 6 5.3: 24, 25, 26
5.4: 29, 30, 32
6.1: 1, 2, 3, 4, 7, 8
Mar 04 - Mar 08
EXAM: Wed, 3/06
Solutions
tables used in the exam 		6, 7 6.2: 11, 12, 13, 14, 15, 16
6.3: 17, 18, 20, 22, 24, 30
7.1: 1, 7, 8
Mar 18 - Mar 22
Homework 6: 3/22
Solutions 		7, 8 7.2: 12, 13, 16
7.3: 17, 18, 19
8.1: 1, 2, 3
Mar 25 - Mar 29
Homework 7: 3/29
Solutions 		8, 9 8.2: 6, 7, 8
8.3: 9, 10, 13, 16
9.1: 5, 6, 8
9.2: 11
Apr 01 - Apr 05
Homework 8: 4/05
Solutions 		9, 10 9.3: 17, 20, 22
10.1: 2
Apr 08 - Apr 12
EXAM: Wed, 4/10
Solutions 		10 10.2: 4, 5, 6, 7, 8
10.3: 9, 10, 11, 12, 13
10.4: 15, 16, 20, 23, 24
Apr 15 - Apr 19
Homework 9: 4/19
Solutions 		10, 11 11.2: 2, 3, 4
11.4: 11, 12, 17
11.5: 22, 23, 28
Apr 22 - Apr 26
Homework 10: 4/26
Solutions 		11, 12 11.6: 30, 34, 37
11.7: 40, 41, 42, 44, 45, 46
12.1: 1, 2, 5, 7
12.2: 10, 12
12.4: 21, 22, 28
May 2
FINAL EXAMINATION
(3:30 PM - 5:30 PM)
Solutions

Help Resources

There are a number of ways to get help with hard problems or for understanding some material a little better:

Office Hours

My office hours are above. You should feel free to come in during those times without appointment. If you can't make it during those hours, you may call me or send me an email message and, with luck, we may be able to find another time.

The help desk at Cornerstone

Stats Help hours (including the hours for Professor and TA Office Hours, and the Walk-In Help on the South 40) will all be listed on the google calendar at Cornerstone: http://cornerstone.wustl.edu/Home/CornerstoneCalendar.aspx.

Walk-In Help for Statistics will begin later this month and will be held in the Cornerstone Tech Lab in Gregg Hall on the South 40. Check the Cornerstone Calendar for times and locations.

Questions about academic resources for stats students? Contact Lisa Kuehne at lmkuehne@wustl.edu.

Study Groups and Appointments with Mentors can also be arranged by visiting Cornerstone (in Gregg Hall) during regular business hours to talk with an advisor about your needs.

Renato Feres
feres@math.wustl.edu
Cupples I, Room 17
(314) 935 - 6752 (office phone)

Department of Mathematics
Washington University
Campus Box 1146
Saint Louis
Missouri, 63130 USA.




Last Updated: January 10, 2011