Longitudinal Data Analysis   A  2-Day Course on Regression Methods for Analyzing Panel Data 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.  There are four widely available methods for dealing with dependence:  robust standard errors, generalized estimating equations, random effects models and fixed effects models.  This course examines each of these methods in some detail, with an eye to discerning their relative advantages and disadvantages.  Different methods are considered for quantitative outcomes, categorical outcomes, and count data outcomes.     This course is based in part on Paul Allison’s Fixed Effects Regression Methods for Longitudinal Data Using SAS, published by the SAS Institute in 2005, and his Fixed Effects Regression Models, published by Sage in 2009.   Who should attend?   If you need to analyze longitudinal data and have a basic statistical background, this course is for you. You should have a good working knowledge of the principles and practice of multiple regression, as well as elementary statistical inference. And it is also helpful to have some familiarity with logistic regression. But you do not need to know matrix algebra, calculus, or likelihood theory.    Materials   Participants receive a bound manual containing detailed lecture notes (with equations and graphics), examples of computer printout, and many other useful features. This book frees participants from the distracting task of note taking.   Registration and lodging   The fee of $750 includes all course materials and a continental breakfast each day.  If you register before March 1, the fee is discounted to $650. To ensure your participation, click here to register. Cancellations received two weeks before the course begins are fully refundable (minus a $50 processing fee if you paid by credit card).   Participants must make their own arrangements for lodging and meals. Special reduced rates have been arranged at a nearby hotel.   Computing   This course will use SAS for the many empirical examples, but lecture notes using Stata are also available to course participants.  No computers will be provided on site and there will be no supervised exercises.  However, you are welcome to bring your own laptop and perform the distributed exercises on your own time.     Course outline Opportunities and challenges of panel data. Data requirements Control for unobservables Determining causal order Problem of dependence Software considerations Linear models Robust standard errors Random effects models Fixed effects models Hybrid models Logistic regression models Robust standard errors Subject-specific vs. population averaged methods Random effects models Fixed effects models Hybrid models Count data models Poisson models Negative binomial models Fixed and random effects Linear structural equation models Fixed and random effects in the SEM context Models for reciprocal causation with lagged effects Comments from recent participants   Participants in two courses were asked to rate the course on a 10-point scale. Of the 73 who responded, the mean rating was 8.7. They were also invited to write attributed comments about the course. Here are all the comments that were received:   "A very well organized course that is extremely useful for anyone interested in applying longitudinal data analysis methods in academic research."   Tatiana Manolova, Bentley University   "Paul Allison is the best in the business. After years of looking things up in his books, his longitudinal data analysis class exceeded my expectations. I am eager to put my new skills to work." Emily Parker, Health Partners Research Foundation   "This was one of the best stats courses I've every taken, the other being Allison's Missing Data course.  I understood concepts that I didn't grasp through my previous stats courses. The information is very clearly explained, and practical examples used to illustrate concepts. The rationale behind why things are done in a certain way is clearly explained." Margaret Hsieh, University of Pittsburgh Medical Center   "The literature on repeated measures and longitudinal data is immense because there are numerous models which can be used to analyze the data. Dr. Allison's Longitudinal Data Analysis course provides careful guidance on these models, the differences between them, and the inferences that can be made from the results. It provides the student with confidence to use the models and forms a basis for additional reading." Robert Goldberg-Alberts, Wyeth   "A very hands-on course taught by a very knowledgeable, kind and helpful professor.  Highly recommended." Enya He, University of North Texas   "Thorough, clearly presented treatment of a complex topic." Palmer Bessey, Weil Cornell Medical College   "I had taught myself some longitudinal methods, such as GEE. What I lacked was a good understanding of the relative pros and cons of each model. Dr. Allison's structured and iterative discussion of robust standard errors, GEE, random effects, fixed effects and hybrid methods helped to describe each method in comparison with the others. This will help me in making model fit decisions in my research." Matt Epperson, Rutgers University   "This a great course for those with some exposure to correlated data, but are unsure how to best work with their data. Very useful combination of statistical theory and hand-on application of SAS programming."   Kate Bauer, University of Minnesota   "This is a good course to stretch your statistical knowledge. I will definitely recommend it as a challenge to those looking for a refresher course on longitudinal analysis." Elizabeth Vasquez, Helen Hayes Hospital   "Dr. Allison's deep understanding of this topic and his ability to teach it is outstanding. This course was a great opportunity to learn from the #1 expert in longitudinal data analysis for panel data." Michelle Lalonde, University of Toronto   "This is a very good training course for panel/longitudinal data analysis, which only takes two days to learn both linear and non-linear analysis. Very good course arrangement and coding examples. Highly recommended!" Jianjun Zhang, West Virginia University   "Professor Allison is an excellent teacher! I learned a lot from this class. I would highly recommend this class to anyone who does longitudinal data analysis." Chang Liu, Merck   "Very helpful class! I would recommend it to anybody trying to understand various methods for analyzing longitudinal data." Janet Grubber, Duke University Medical Center