ICPSR Summer Workshop on Categaorical Data Analysis
July 8-12, 2013 | Instructor Scott Long | University of Michigan-Ann Arbor
To enroll, contact the ICPSR Summer Program · FAQs ·
What to wear ·
Getting started with Stata ·
Lab files ·
Lecture files · Math Review ·Books
The workshop deals with the most important regression models for binary, ordinal, nominal and count outcomes. While advances in software make it simple to estimate these models, the effective interpretation of these nonlinear models is a vexingly difficult art that requires time, practice, and a firm grounding in the goals of your analysis and the characteristics of your model. The workshop begins by discussing the general objectives for interpreting results from any regression model and considers why these objectives are more difficult in nonlinear models. Concepts of estimation, testing, and identification are introduced in a quick review of the linear regression model. These ideas are used to develop the binary logit and probit models. Advanced methods of interpretation are introduced using Stata's margins command, with detailed examples on how to compute and interpreting average marginal effects, the distribution of effects, and related methods. Concepts from the binary model are used to develop the multinomial logit model for nominal outcomes, followed by the development of several models for ordinal outcomes. Finally, models for count data, including Poisson regression, negative binomial regression, and zero modified models are presented. While familiarity with Stata is recommended, the labs provide step by step instructions for those new to Stata.
Here are some highlights of the class.
- "Predictive margins" involve interpreting models using predictions rather than parameters. This provides incredibly valuable insights into your models.
- Mastering the powerful (and confusing) margins command using new SPost commands to control margins.
- Automating your work to make complex analyses routine.
- Using graphical methods for exploring models.
- Finding out why "ordinal" models aren't always ordinal and why ordinal outcomes might not be ordinal.
While the labs use Stata and the lectures illustrate concepts with analyses from Stata, the class does not assume you have used Stata. The labs provide step by step instructions to help you learn Stata quickly. More experience Stata users can jump into more advanced exercises. .
- Can I bring my own data? Yes and I am pleased to meet with
you individually to discuss your research. It is important that the data has been "cleaned" since you won't have time to do that during the workshop. Ideally, data should be converted to Stata format. Software is available to
convert data from one format to another, but to be safe try to do this
- Do I have to use Stata?
If you don't want to try to use Stata at all, you probably won't want to take this course. If you have never used Stata but are willing to give it a try, you will have no
trouble doing the exercises. In lab you will be given handouts that walk you
through each step and the TA and I will be there to help. Will you be able
to apply what is done using other software? Yes, but but it is likely to
require a great deal more work. Why? Jeremy Freese and I have written the SPost commands that make complex computations
am glad to talk with people about how you might approach these
computations in other software.
- Do I need to know everything on
the math review? It helps, but the most important thing is to be
sure you are comfortable with the log transformation and the exponential. If
you are confused by these, be sure to ask when you get here.
- What should I bring (to wear)? A sweater or
light jacket for class! While the thermometer in the class room is a
continuous scale, it tends to produce a binary outcome of being either too
hot or too cold. This might have changed, but to be safe, bring something
extra to wear. (Those of you who have been reading ahead will notice that
I've just given an example of the latent variable approach to deriving the
binary logit and probit models.)
- Should I bring a laptop? In the past many participants have used their laptops for lab exercises. We can installed a Stata with a temporary license for use during the workshop. S
- And, what
else should I bring? I recommend a USB storage device to save your work.
- Will I have fun? I
certainly will and I think you will too. ICPSR workshops are a
great way to learn new methods.
- Do I need to buy the books? You will be given extensive handouts with lecture notes.
While at ICPSR this might be all you need. The "with Stata"
book has less technical detail than the Sage book, but has a lot of
information of using Stata.
J. Scott Long, 1997, Regression Models for Categorical and Limited Dependent Variables. Advanced Quantitative Techniques in the Social Sciences, Volume 7. Sage Publications.. This books has a more technical discussion of topics in the class.
J. Scott Long and Jeremy Freese, 2005, Regression Models for Categorical Dependent Variables Using Stata, 2nd Edition. Stata Press: College Stata, TX. This book does not use margins, which is highlighted in class.
Long, J.S. 2008, The Workflow of Data Analysis Using Stata. Stata Press: College Station, TX. If you plan to do a lot of data analysis, this book will save you a lot of time and make your work replicable.
Datasets and sample do files
In Stata you should be able to get to the datasets over the
web using spex dataset-name. The do files
and data can be downloaded from Stata by type findit icpsrlong and following what it tells you.