The course aims to treat the extensions of the linear model for the analysis of complex data such as mixed linear models (LMM), generalized linear models (GLM), in particular the log-modelPoisson’s linear counting data and binomial logistics model. Case analysis is carried out in an R environment. Knowledge and understanding. This teaching will provide knowledge and understanding skills relating to: 1) Structure of the main linear models in the presence of continuous and discrete categorical variables, both on the dependent variable and on the independent variable; 2) Methods for selecting and defining the model in relation to the nature of the data; 3) Main procedures implemented in R for the construction of linear models and the subsequent deepening of the analyses with the relative graphic representations 4) Reading and interpreting the outputs of the analyses produced with R.
Course Prerequisites
knowledge of the topics covered in the basic courses of Statistics (distributions: Gauss, Pisson and Binomial, Linear models). Knowledge of R.
Teaching Methods
Theoretical lessons in the classroom and practical exercises with the software R.
Assessment Methods
The exam is written and consists in the preparation of a statistical analysis of data with the software R (test duration: 2 hours), which deals with topics of both theoretical and practical nature.
Texts
Zuur et al. Mixed effects models and extensions in ecology with R
Contents
Application of linear models to biological data analysis: 1) Extension of linear models; mixed linear model (LMM) and generalized linear model (GLM) theory: model specification, estimable functions, testable hypotheses. Parameterization of effects and reference category. Contrasts 2) Analysis of data sets for each model with R 3) General linear mixed models (GLMM)
