Introduction to Linear Mixed-Effects Models, GLMM and Multivariate GLMM with R

Live online (and occasionally onsite) course :

Introduction to Linear Mixed-Effects Models, GLMM and Multivariate GLMM with R

 

Zoom sessions are scheduled for: 14.00 - 19.00 UK time; 28 and 30 November, 1, 4, 5, 6, and 7 December 2023. Another live course will be scheduled in 2024.

 

The course begins with a brief review of multiple linear regression and generalized linear models. This is followed by an introduction to linear mixed-effects models and generalized linear mixed-effects models (GLMM) for analyzing hierarchical or clustered data. Examples of such data include multiple observations from the same animal, site, area, nest, patient, hospital, vessel, lake, hive, transect, and so forth. These statistical techniques are designed to address dependency within your data.

In the second part of the course, we apply GLMMs to various types of data. This includes continuous data (e.g., biomass), binary data (e.g., the presence or absence of a disease), proportional data (e.g., % coverage), and count data. For these analyses, we employ several distributions: Poisson, negative binomial, Bernoulli, binomial, beta, ordered beta, Tweedie, and gamma.

In the final part of the course, we delve into multivariate GLMMs, especially generalised linear latent variable models (GLLVM). These models allow for the analysis of multiple response variables within a single model using the 'gllvm' package.

The course website contains the following modules.

Preparation material (containing on-demand video):

  • Revision exercise on multiple linear regression.
  • Introduction to matrix notation.

Module 1

  • General introduction.
  • Theory presentation for linear mixed-effects models for nested data.
  • Two exercises on linear mixed-effects models with random intercepts.
  • Exercise showing how to apply a two-way nested linear mixed-effects model.

Module 2

  • Exercise on linear mixed-effects models with random intercepts and slopes.
  • Revision exercise Poisson GLM and negative binomial GLM.
  • Introduction to DHARMa.
  • Exercise on Poisson GLMM.

Module 3

  • Exercise on Negative binomial GLMM.
  • Exercise on two-way nested and crossed random effects in a GLMM.
  • Exercise showing how to apply a Bernoulli GLMM for the analysis of absence-presence data.
  • Exercise showing how to apply a binomial GLMM for the analysis of proportional data.

Module 4

  • Exercise showing how to apply a beta GLMM for the analysis of coverage data.
  • Exercise showing how to apply an ordered beta GLMM for the analysis of coverage data that includes zeros and/or ones.
  • Exercise showing how to apply a gamma GLMM for the analysis of continuous positive data.
  • Exercise showing how to apply a Tweedie GLMM for the analysis of continuous positive data (with zeros).

Module 5

  • Catching up.
  • Theory presentation on multivariate GLMM.
  • Three exercises showing the application of multivariate GLMM using the gllvm package.

We will predominately use the R package glmmTMB for the GLMM exercises. Multivariate GLMMs will be fitted using gllvm.

The course material consists of relevant pdf files of presentations, data sets, and clearly documented R code. Course participants will be given access to the course website with all data sets, R solution code, and course material 1 week before the start of the course.

 

PRE-REQUIRED KNOWLEDGE:
Working knowledge of R, data exploration, linear regression and GLM (Poisson, negative binomial, Bernoulli). This is a non-technical course.

 

FREE 2-HOUR FACE-TO-FACE MEETING
Provided that you attend 80% of the Zoom sessions, the course fee includes a 2-hour face-to-face meeting with one or both instructors. You can discuss your own data, but we strongly advise that the statistical topics are within the content of the course. The 2-hour consultancy needs to be consumed in two 1-hour sessions, and will take place at a mutually convenient time. It is not transferable. The meetings need to take place within 12 months after the last live Zoom module. If attendance is between 50-80%, we will offer you 1 hour face-to-face consultancy. If attendance is less than 50%, we will not offer face-to-face consultancy.

This course does not contain on-demand video.