Professor Dan Myers
Office: 735 Flanner Hall, 631-3839
Myers.33@nd.edu
Office Hours: by appointment
Web Page for this course: http://www.nd.edu/~dmyers/courses/593sp01/
 
TA: Yan (Elaine) Li, yli@nd.edu
 
Purpose of the Course:
 
This course is a follow up to the introductory statistics course, Soc 592, that most of you took last semester. In the present course, we will begin where you left off last semester. We will first briefly review what you have learned about ordinary least-squares regression models. The remainder of the course will extend OLS and introduce techniques that allow the estimation of multivariate models when the assumptions of OLS are not met.
 
Required Texts:
1. Agresti, Alan and Barbara Finlay. 1997. Statistical Methods for the Social Sciences. Prentice Hall. (If you took 592, you should have this book already. If you did not, do NOT buy it. We will be using a relatively small portion which you can copy from one of the other students.)
2. Allison, Paul D. 1999. Multiple Regression: A Primer. Pine Forge Press.
3. Hamilton, Lawrence C. 1998. Statistics with Stata 5. Duxbury.
4. Berry, William D. and Stanley Feldman. 1985. Multiple Regression in Practice. Sage.
5. Fox, John. 1991. Regression Diagnostics. Sage.
6. Jaccard, James, Robert Turrisi, and Choi K. Wan. 1990. Interaction Effects in Multiple Regression. Sage.
7. Menard, Scott. 1995. Applied Logistic Regression Analysis. Sage.
8. Allison, Paul D. 1984. Event History Analysis. Sage.
9. Berry, William D. 1984. Non-Recursive Causal Models. Sage.
10. Packet of course handouts. Available from Nancy on the 8th Floor of Flanner.
 

Course Requirements:   1. 2 Exams: 15% each (3/7/01 and 4/11/00) 2. 1 Term Project: 40% (Due 5/4/00) 3. Homework assignments: 30%

Grade Scale:   A: 93-100% A-: 90-92 B+: 87-89, B: 83-86, B+: 80-82 C+: 77-79, C: 73-76, C+: 70-72 D: 60-69, F: Below 60%

 
 
COURSE OUTLINE
 
I. OLS Review and Interactions
 
Jan. 17, 22, 24: Review OLS, Dummy Coding and Familiarize with Stata
Review A & F Ch. 9, 10, 11.0-11.4
Allison Preface, Ch. 1-2, 5, skim Ch. 4,
Hamilton Ch 1-4; Ch. 6 (pp. 129-154 only)
 
Jan. 29, 31, Feb. 5: Interaction in ANOVA and Regression
A & F 12.1-12.5, 11.5
Jaccard et al. Ch. 1-3.
Hamilton pp. 114-126
 
II. Assumptions and Problems
 
Feb. 7: Measurement Error
Allison Ch. 3, Ch. 6
B & F Ch. 3
 
Feb. 12: Model Mis-specification
B & F Ch. 2
 
Feb. 14, 19: Nonlinear/Curvilinear
Allison Ch. 8
A & F 14.4, 14.6
B & F Ch.5
Hamilton pp. 157-160, 193-197
Jaccard et al. Ch. 4.
 
Feb. 21: Regression Diagnostics and Heteroskedasticity
Review Allison Ch. 3, 6
Skim all of Fox
Fox Ch. 6
B & F Ch. 6
 
***PAPER PROPOSALS DUE: February 21***
 
Feb. 26: Outliers
Fox Ch.4
Hamilton Ch. 7
 
 
 
 
Feb. 28: Multicollinearity
Allison Ch. 7
B & F Ch. 4
Fox Ch. 3

***EXAM I: MARCH 5***
 
Mar. 7: Missing Data (NO LAB ON MARCH 9)
 
SPRING BREAK
 
Mar. 19: Supressors
 
III. Other Regression Type Models
 
Mar. 21, 26: Event Count Models (Poisson and Negative Binomial Regression)
Hamilton pp. 271-275
 
Mar. 28, Apr 2, 4, 9: Logistic and Log-linear models
A & F 15
Menard
Hamilton pp. 225-241
 
***EXAM II: APRIL 11***
 
Apr. 18, 23: Survival Time Models
A & F 16.1
Allison (Sage)
Hamilton pp. 250-271
 
IV. Causal Modeling
 
April 25, 30, May 2: Path Analysis and Non-recursive Models
A & F 16.2, 16.4
Berry
 
***TERM PROJECTS DUE MAY 4 at 4:00 pm***
 
 
Policies/Advice
 
Experience leads me to believe that most of you will find this to be a challenging course. For this reason, I want to encourage you to be very systematic in your work for this course. This means:
 
1. Attend class. Attending is critical. It is unlikely you will be able to absorb all the material simply by reading the books. Much confusion can be eliminated by attending and participating.
 
2. Reading. Reading assignments are absolutely essential in this course. You must do the assigned reading prior to coming to class each time. I will make a lot more sense to you and will be able to help you much more if you have done the reading in advance.
 
Most of you will find that understanding the material requires going over it more than once in some way--that is simply the nature of the beast when it comes to statistics. Particularly if you anticipate having trouble in this course you might consider reading the material more than once (is this guy nuts?) or even more extreme, actually taking notes as you read (yep, he's definitely lost it).
 
3. Keep up on your work--don't fall behind. This course builds as it goes and you will find yourself becoming very confused if you have gaps you have missed. At this level, I really don't think testing with exams is all that critical, but they do force you to keep up and so they are a part of this course.
 
4. Working together. Graduate study should be a collective rather than an individualistic or competitive process. Therefore, I encourage you to work closely with others in the class. By the end of the course, I will be happiest if you understand the material--no matter how you got there. That said, you should still do your own final write-up of your homework and papers. You should also avoid becoming dependent on those who find this material easier to grasp. The point is to learn as much as you can, not to just get it done without really understanding what you are doing.
 
5. Keep in close contact with your TA. She is here to help you. This is her main responsibility and you should take advantage of this resource.
6. Papers and homework assignments are due at the start of the class on the due date. Unless you have made prior arrangements with me, no assignments will be accepted after that point. This policy exists to keep you from missing class while putting the final touches on you work. Plan ahead so you won't have this problem.
 
7. Exams cannot be made up nor will I issue the grade of incomplete. The only exceptions are in the case of serious illness or a death in the family.
 
Term Project
 
What I consider to be the most important requirement of this course is an empirical research paper that uses some of the methods covered in this course to examine a sociological substantive problem. The expected output for this assignment is something like an empirical paper you would see in a sociology journal.
 
I am not particularly interested in the length of your paper, but it should be a journal length treatment and as a rule-of thumb, I would not be surprised if the average paper were over 20 pages in length. You paper should include the formulation of a research problem, a linkage to the previous substantive literature, an analysis of pertinent data, and a discussion of findings. If you handle each of these tasks completely, the length of your paper should not concern you in the slightest.
 
Because this is a statistics class, you may be tempted to slight the literature review and problem formulation components of your effort in favor of endless methodological and statistical detail. Resist this temptation because it will not serve you well. Statistical bluster at the expense of meaning and contribution to the literature is a bad mistake.
 
To help you avoid this problem, I will require you to write a paper proposal early in the course, before we have covered much of the statistical material you will need to use. This way, you can concentrate on substance prior to running models with your data. The proposal draft is due on February 21 at the start of the class. If you would like to see an example of a proposal that I would consider acceptable, I have placed an example of one I wrote for a similar class on the course web page.
 
Prior to writing your proposal, you should be certain that data are at your disposal which can be used to examine the question you have in mind. The Lab for Social Research has many data sets in its possession that are freely available to you. There are also tons of data available on the web these days. On my personal web page, there is a link to Data on the Web. This is just a small list to get you started in finding data--there are hundreds of other sources. Faculty who work in your areas of interests may also have data that they will allow you to access. If you are planning on a quantitative MA thesis, this is a good chance to get a jump on it and possibly even produce a preliminary draft.
 
The due date for the project is May 4 at 4:00 pm. Late papers will not be accepted.