Spring 2012
NOTE: These are the Spring 2012 course notes for the second semester of my graduate statistics courses. The notes for the first semester course, Sociology 63992, are also available. These pages make extensive use of Stata and SPSS. If you are mostly interested in learning how to use Stata, the Stata Highlights page lists several of the most relevant handouts from both courses. Some pages are more "stand alone" than others, so adjacent handouts may help clear up any questions you have.
These pages will be updated whenever I complete another session of the course, and possibly sooner. Notes from the current semester when the course is being taught can be found here.
Feel free to email Richard Williams if you have comments or suggestions.
The following special types of files are used on this web page:
Pdf files. Require Adobe Acrobat. Toolbook files. SPSSWIN files. Necessary for doing homework problems. Can probably be adapted for other SPSS platforms. You should save these files to your local hard disk and then use them with SPSS. Stata files. Necessary for doing homework problems.You should save these files to your local hard disk and then use them with Stata. In addition, some files are in zipped (compressed) format. If you don't have an unzipping program (e.g. Winzip), you can use the free PC Magazine PCDEZIP utility.
Finally, please note that the answer keys for the exams and homework differ in the amount of detail provided. I sometimes give very detailed answers, other times the answers are much more minimal (and given the information provided I assume the student can figure out the rest). Students should always aim for complete answers in their homework and exams. In particular, it is hard to give partial credit when it is not clear why an error was made.
Readings Packet (You need a Notre Dame NETID to access these)
Notre Dame's Center for Social Research (has links to several data sets and describes support services)
Dropbox. I strongly encourage you to set up a Dropbox account if you do not already have one. Dropbox gives you a minimum of 2GB of free online storage. More critically, with Dropbox you can set up shared folders. This makes it much easier when you want me or others to help you with your research. You can create a folder, put your data and programs in it, and then share the folder with me. If you set up an account use your .edu email address because you can get more bonus storage that way. For more click on this link.
Useful sites for learning about Stata and SPSS
UCLA's Statistical Computing Resources RW Suggestions for Using Stata at Notre Dame UCLA's SPSS Starter Kit Resources for learning Stata UCLA - How does Stata compare with SAS and SPSS? The Stata User Support Page Ben Jann's estout/esttab support page (esttab & estout are great for formatting output from Stata)
PART I: In this section, we briefly review the basics of OLS regression. We talk about some of the most common issues (measurement error, missing data, violations of OLS assumptions) encountered in regression analysis.
reg01.sav - Data file used in the SPSS Regression handout
reg01.dta - Data file used in the Stata Regression handout
Also Recommended: Robert Yaffee's Robust Regression Modeling with Stata (This is 93 pages long but it is basically overhead slides and hence much shorter than it at first appears to be. Nice discussions of how to deal with outliers and with heteroskedasticity.)
Also Recommended: UCLA's Regression Diagnostics Page. Shows a lot of the techniques that are available with Stata for detecting outliers, heteroskedasticity, multicollinearity, serial correlation and other problems with regression models.
PART II: This section shows how regression can be used to properly specify a causal model. We begin by introducing "the logic of causal order," which lets us understand the different kinds of causal relationships that might be present between variables. Common model mis-specifications are then addressed (e.g. omitted variables, extraneous variables, variables with nonlinear effects). We discuss how to choose between alternative causal models. Finally, we introduce path analysis as a method for causal modeling.
tbklogic.zip These are toolbook presentations which we will go over in class.
[NOTE: You may have trouble using this with Win 7. I did with Win7 64 bit. But it can be done. If you really want to do so I can tell you how to do it.]
PART III: Here, we develop path analysis techniques more fully. We talk about more complicated models that cannot be accurately estimated through conventional OLS regression techniques (e.g. nonrecursive models). We also talk about situations where the nature of the data make OLS regression inappropriate (e.g. dichotomous dependent variables) or less than optimal.
Brief Overview of Survey Data Analysis. By default, most statistical techniques assume that data were collected via simple random sampling. This is often not true for large national data sets. Fortunately, Stata makes it easy to analyze such data, but there are some important differences in how you go about testing hypotheses and assessing model fit.
Introduction to
Survey Data Analysis (From the Stata 12 documentation; read the first
few pages carefully and skim the rest)
Analyzing Survey Data: Some Key Issues to be Aware of
UCLA's (see lower third of page) and StataCorp's FAQS on Survey Data Analysis (Optional; you may want to refer to these if you use the SVY commands)
Brief Overview of Other Advanced Methods.
Extremely Brief Overviews of Event History Analysis and Hierarchical Linear Modeling --
Read Ch. 9 of Paul Allison's Multiple Regression Primer, paying particular attention to section 9.9 (Multilevel Models) and section 9.12 (Event History Analysis)
Advanced Categorical Data Methods. We will only cover these in class if time permits, but skim through them on your own. These are covered in much more depth in Soc 73994, Categorical Data Analysis.
Other materials may be available upon request.
