Kaisa_2012_3_photo by Veikko Somerpuro

12.8.2019 at 09:00 - 18.9.2019 at 15:00


Here is the course’s teaching schedule. Check the description for possible other schedules.

Mon 11.11.2019
09:00 - 16:00
Tue 12.11.2019
09:00 - 16:00
Wed 13.11.2019
09:00 - 16:00
Thu 14.11.2019
09:00 - 16:00
Fri 15.11.2019
09:00 - 16:00


Participants must have taken YEB doctoral school's YEB-116 Data exploration, regression, GLM & GAM course or otherwise acquired the skills taught on that course:

  • Data exploration (outliers, collinearity, transformations, relationships, interactions)
  • Linear regression (model selection, interactions, dealing with categorical covariates, sketching model fit)
  • GLM with various distributions (Poisson GLM, negative binomial GLM, Bernoulli GLM) and dealing with overdispersion
  • GAM with various distributions (Gaussian GAM, Poisson GAM, negative binomial GAM)

Mixed modelling and MCMC is strongly recommended, but a short revision is provided.

Participants must be able to run the above analyses in R.

Each participant's skills will be tested by a simple on-line quizz before the course starts and seats will be offered to doctoral candidates with background skills needed for course completion.

After completing the course the participants recognise pseudo-replication and know how to deal with it as well as with data with non-linear patterns. Furthermore, participants will be familiar with the following methods and concepts:

  • Bayesian statistics
  • MCMC
  • multiple linear regression
  • generalised linear models
  • mixed effects models
  • generalised additive models (GAM)
  • generalised additive mixed effects models (GAMM)

The course unit can be completed at any time during doctoral studies.

The course unit is offered every third year in the autumn term (2019, 2022 etc.)

  • The course begins with a revision of multiple linear regression, generalised linear models, mixed effects modelling and Bayesian statistics (MCMC).
  • Sometimes, parametric models do not quite fit the data and in such cases generalised additive models (GAM; a smoothing technique) can be used.
  • We will explain and illustrate GAMs to analyse continuous data, count data and binary data.
  • In the second part of the course we use generalised additive mixed effects models (GAMM) to analyse nested (also called hierarchical or clustered) data, e.g. multiple observations from the same animal, site, area, nest, patient, hospital, vessel, lake, hive, transect, etc.
  • During the course several case studies are presented, in which the statistical theory is integrated with applied analyses in a clear and understandable manner.
  • We will use frequentist (mgcv, gamm4) and Bayesian tools (MCMC in JAGS).

The structure of the five-day intensive course is built on introductory lectures and demonstrations to each topic coupled with computer exercises run in R. The course is taught by Dr. Alain Zuur and Elena Ieno from Highland Statistics (www.highstat.com).

PDF files of all PowerPoint presentations are provided to course participants. The files are based on various chapters from:

  • Zuur AF, Hilbe JM and Ieno EN: A Beginner’s Guide to GLM and GLMM using MCMC with R. (2013).
  • Zuur AF and Ieno EN: A Beginner’s Guide to Zero Inflated Models with R. (2016).
  • Zuur AF, Saveliev AA, Ieno EN: A Beginner’s Guide to GAM and GAMM with R. (2014).

The course can be followed without purchasing these books.

Grading: pass/fail
As the course is very demanding and advances fast, full attendance is required. Missing even a half day of teaching would make following the rest of the course very hard, if not impossible.

Instruction language: English.

Course teacher