The course will provide an introduction to longitudinal data analysis using mixed effects regression models, drawing on material from the book Longitudinal Data Analysis (Wiley, 2006) using the new SuperMix statistical software program (a 15-day trial edition of SuperMix is available at www.ssicentral.com/supermix/index.html). The focus will be on application of these models, with direct application illustrated using SuperMix.
In particular, the basic mixed-effects regression model for continuous outcomes will be introduced and described, including use of polynomials for expressing change across time, the multilevel representation of the mixed model, treatment of time-invariant and time-varying covariates, and modeling of the variance-covariance structure of the repeated measures.
It will be shown how these models can allow for missing data across time in terms of the outcome variable, thus permitting analysis of subjects who have incomplete data across time.
Finally, because categorical outcomes are common in many research areas, description and application of mixed-effects logistic regression models will also be covered.
Attendees are encouraged to download the trial software onto their laptops prior to course, and to bring their laptops with them to the course.