Use linear, logistic, Poisson and its variants regressions in the context of longitudinal studies in health and medical research.

Western Sydney University presents:

An online short course on Longitudinal Modelling for Health Researchers

Instructor: Dr Haider Mannan, Senior Lecturer in Biostatistics, Western Sydney University

Use linear, logistic, Poisson and its variants regressions in the context of longitudinal studies in health and medical research.

The course will be run over 3 days via Zoom sessions

• DAY 1 - March 21, 2022: 9am - 3pm (6 hours + 1 hour break)
• DAY 2 - March 22, 2022: 9am - 2pm (4 hours + 1 hour break)
• DAY 3 - March 23, 2022: 9am - 2pm (4 hours + 1 hour break)

In this course, you will learn about the commonly used regression models in longitudinal data analysis. The focus will be on multilevel mixed effects and Generalized Estimating Equations models but there will also be examples for fixed and random effects models, hybrid and Mundlak models, the mixed-effects location-scale, conditional logistic, and segmented linear regression models.

Almost all examples will use data from population health research exploring the complexities of designing and implementing the most appropriate statistical model and providing a clear interpretation of results. From a simple longitudinal study design to more complex designs such as nested, crossover, repeated measures for both covariates and outcome, interrupted time series and ecological momentary analysis will be covered.

The course is online with interactive exercises and opportunities for one-on-one support from the instructor. If you have any immediate questions, please contact us via WesternX@westernsydney.edu.au.

The course costs 950 AUD. WSU students and alumni can contact WesternX for a 50% discount! Participants can complete parts of the course, e.g., modules 1 and 2 or modules 1 and 2 plus 3 or 4.

Upon completion of the course participants will receive certificate of completion.

What some of our students said during the course's first run

“Learning many new statistical methods and tools, understanding the statistical way of thinking. The teacher’s dedication and very responsive and helpful attitude helped me to achieve and learn a lot from this course. I would say this course is very informative and learning materials are constructed in a way that really facilitated the learning process.” (Student comment, Longitudinal Modelling for Population Health Researchers, Spring 2020)

“The instructor set very important examples while explaining the theory. His quizzes are very important for students to have a clear idea of the topics.” (Student comment, Longitudinal Modelling for Population Health Researchers, Spring 2020)

“He simplified the Stata programming language which really motivated us to learn it better and hence enhanced our skills in use of Stata. He even offered us to meet after the class for deeper understanding of programming which is beyond the scope of the course. He tried his best to give examples to simplify the statistical methods taught in the course.” (Personal Communication, Longitudinal Modelling for Population Health Researchers, Spring 2020)

Course Outcomes are:

·         Use linear, logistic, Poisson and its variants regressions in the context of longitudinal studies in health and medical research.

·         Identify advantages of longitudinal study design; multilevel data structures for longitudinal studies; assumptions, pros and cons of multilevel & GEE models; their common correlation structures; sample size & number of waves for fitting them.

·         Fit and interpret multilevel linear models under simple and complex longitudinal study designs such as nested, crossed effects, and repeated measures studies with time-varying covariates.

·         Fit and interpret multilevel logistic, Poisson and its variants regressions under longitudinal study designs; segmented linear model under interrupted time series and the location-scale model under ecological momentary analysis.

·         Fit and interpret GEE linear, logistic, Poisson and its variants regressions under various longitudinal study designs.

The course is split into four separate modules:

1. What are fixed, random, and mixed-effects, hybrid, and Mundlak models for analyzing longitudinal data for a continuous outcome?
2. Applications of fixed, random, and mixed-effects, hybrid, and Mundlak models for analyzing longitudinal data in population health for a continuous outcome.
3. Methods and applications of Generalized Linear Mixed and conditional logistic models for analyzing longitudinal data in health for a dichotomous outcome.
4. Methods and applications of GEE, the mixed-effects location-scale, and segmented regression models for analyzing longitudinal data in health.