Longitudinal Data Analysis Using SAS - Online Course
A 4-Week On-Demand Seminar Taught by
Paul AllisonNo upcoming dates
For many years, Dr. Paul Allison has been teaching his acclaimed two-day seminar on Longitudinal Data Analysis Using SAS to audiences around the world. This course covers several popular methods for the analysis of longitudinal data with repeated measures: robust standard errors, generalized least squares, generalized estimating equations, random effects models and fixed effects models.
The course takes place in a series of four weekly installments of videos, quizzes, readings, and assignments, and requires about 4-6 hours/week. You can participate at your own convenience; there are no set times when you are required to be online. The course can be accessed with any recent web browser on almost any platform, including iPhone, iPad, and Android devices. It consists of 10 video modules:
- Advantages and disadvantages of panel data
- Robust standard errors and generalized least squares for linear models
- Random effects (or mixed) linear models
- Fixed effects linear models
- The between-within method
- Logistic regression with robust standard errors and GEE
- Logistic regression with random effects
- Fixed effects logistic regression
- Models and methods for count data
- Linear structural equation models
Each module is followed by a short multiple-choice quiz to test your knowledge. There are also weekly exercises that ask you to apply what you’ve learned to a real data set.
Each week, there are assigned articles to read. There is also an online discussion forum where you can post questions or comments about any aspect of the course. All questions will be promptly answered by Dr. Allison.
Downloadable course materials include the following pdf files:
- All slides displayed in the videos.
- Exercises for each week.
- Readings for each week.
- Computer code for all exercises (in SAS, Stata, and R formats).
- A certificate of completion.
More details about course content
PANEL DATA OFFER MAJOR OPPORTUNITIES AND SERIOUS PITFALLS
The most common type of longitudinal data is panel data, consisting of measurements of predictor and response variables at two or more points in time for many individuals. Such data have two major attractions: the ability to control for unobservables, and the determination of causal ordering.
However, there is also a major difficulty with panel data: repeated observations are typically correlated, and this invalidates the usual assumption that observations are independent. As a result, confidence intervals and p-values can be severely biased. In some cases, coefficients may also be biased downward.
You’ll learn how to use these methods (robust standard errors, generalized estimating equations, random effects models and fixed effects models) for quantitative outcomes, categorical outcomes, and count data outcomes. You’ll also learn which methods are best suited for different kinds of applications.
This is a hands-on seminar with ample opportunities to practice these new methods.
Here are a few of the topics you won’t want to miss:
- How to use panel data to control for unobserved variables.
- Why fixed effects methods often give very different results from random effects methods.
- How to reshape data from long form to wide form and back again.
- Why the default correlation structure for GEE is usually not the best.
- The difference between maximum likelihood and restricted maximum likelihood.
- How to estimate and interpret random coefficient models.
- Why first-order autoregressive structures are usually unsatisfactory.
- The difference between subject-specific coefficients and population-averaged coefficients, and why it matters.
- How to do longitudinal analysis using ordered logit or multinomial logit.
In this seminar, we will use the following SAS procedures: GLM, SURVEYREG, GENMOD, MIXED, LOGISTIC, SURVEYLOGISTIC, GLIMMIX, and CALIS. Lecture notes using Stata and R are also available.
PANEL DATA OFFER MAJOR OPPORTUNITIES AND SERIOUS PITFALLS
The most common type of longitudinal data is panel data, consisting of measurements of predictor and response variables at two or more points in time for many individuals. Such data have two major attractions: the ability to control for unobservables, and the determination of causal ordering.
However, there is also a major difficulty with panel data: repeated observations are typically correlated, and this invalidates the usual assumption that observations are independent. As a result, confidence intervals and p-values can be severely biased. In some cases, coefficients may also be biased downward.
You’ll learn how to use these methods (robust standard errors, generalized estimating equations, random effects models and fixed effects models) for quantitative outcomes, categorical outcomes, and count data outcomes. You’ll also learn which methods are best suited for different kinds of applications.
This is a hands-on seminar with ample opportunities to practice these new methods.
Here are a few of the topics you won’t want to miss:
- How to use panel data to control for unobserved variables.
- Why fixed effects methods often give very different results from random effects methods.
- How to reshape data from long form to wide form and back again.
- Why the default correlation structure for GEE is usually not the best.
- The difference between maximum likelihood and restricted maximum likelihood.
- How to estimate and interpret random coefficient models.
- Why first-order autoregressive structures are usually unsatisfactory.
- The difference between subject-specific coefficients and population-averaged coefficients, and why it matters.
- How to do longitudinal analysis using ordered logit or multinomial logit.
In this seminar, we will use the following SAS procedures: GLM, SURVEYREG, GENMOD, MIXED, LOGISTIC, SURVEYLOGISTIC, GLIMMIX, and CALIS. Lecture notes using Stata and R are also available.
Computing
This seminar will use SAS for the many empirical examples and the exercises. However, lecture notes and exercises using Stata and R are also available. To do the exercises, you should have a recent version of SAS (or Stata or R) installed on your computer.
There is now a free version of SAS, called the SAS University Edition, that is available to anyone. It has everything needed to run the exercises in this course, and it will run on Windows, Mac or Linux computers. However, you do need a 64-bit machine with at least 1 GB of RAM. You also have to download and install virtualization software that is available free from third-party vendors. The SAS Studio interface runs in your browser, but you do not have to be connected to the Internet. The download and installation are a bit complicated, but well worth the time and effort.
This seminar will use SAS for the many empirical examples and the exercises. However, lecture notes and exercises using Stata and R are also available. To do the exercises, you should have a recent version of SAS (or Stata or R) installed on your computer.
There is now a free version of SAS, called the SAS University Edition, that is available to anyone. It has everything needed to run the exercises in this course, and it will run on Windows, Mac or Linux computers. However, you do need a 64-bit machine with at least 1 GB of RAM. You also have to download and install virtualization software that is available free from third-party vendors. The SAS Studio interface runs in your browser, but you do not have to be connected to the Internet. The download and installation are a bit complicated, but well worth the time and effort.
Seminar outline
- Opportunities and challenges of panel data.
a. Data requirements
b. Benefits of panel data
c. Problem of dependence
d. Software considerations
- Linear models
a. Robust standard errors
b. Generalized least squares
c. Random effects models
d. Fixed effects models
e. Between-within models
- Logistic regression models
a. Robust standard errors
b. Generalized estimating equations
c. Subject-specific vs. population averaged methods
d. Random effects models
e. Fixed effects models
f. Between-within models
- Models for count data
a. Poisson vs. negative binomial models
b. GEE and random effects
c. Fixed effects and between-within models
- Linear structural equation models
a. Fixed and random effects in the SEM context
b. Models for reciprocal causation with lagged effects
- Opportunities and challenges of panel data.
a. Data requirements
b. Benefits of panel data
c. Problem of dependence
d. Software considerations - Linear models
a. Robust standard errors
b. Generalized least squares
c. Random effects models
d. Fixed effects models
e. Between-within models - Logistic regression models
a. Robust standard errors
b. Generalized estimating equations
c. Subject-specific vs. population averaged methods
d. Random effects models
e. Fixed effects models
f. Between-within models - Models for count data
a. Poisson vs. negative binomial models
b. GEE and random effects
c. Fixed effects and between-within models - Linear structural equation models
a. Fixed and random effects in the SEM context
b. Models for reciprocal causation with lagged effects
Who should register?
If you need to analyze longitudinal data and have a basic statistical background, this seminar is for you. You should have a good working knowledge of the principles and practice of multiple regression, as well as elementary statistical inference. It is also helpful to have some familiarity with logistic regression. But you do not need to know matrix algebra, calculus, or likelihood theory.
If you need to analyze longitudinal data and have a basic statistical background, this seminar is for you. You should have a good working knowledge of the principles and practice of multiple regression, as well as elementary statistical inference. It is also helpful to have some familiarity with logistic regression. But you do not need to know matrix algebra, calculus, or likelihood theory.
Payment information
The fee of $995 includes all course materials. The early registration fee of $895 is available until April 28.
PayPal and all major credit cards are accepted.
Group discount rates are available for this course. All inquiries can be sent to info@statisticalhorizons.com.
Our Tax ID number is 26-4576270.
The fee of $995 includes all course materials. The early registration fee of $895 is available until April 28.
PayPal and all major credit cards are accepted.
Group discount rates are available for this course. All inquiries can be sent to info@statisticalhorizons.com.
Our Tax ID number is 26-4576270.