Statistical Analyses with R

Module I. Programming in R and Rstudio

• Along with this model, we will review the most commonly used files in any statistical programming workflow like the scripts (.R) the allocated memory (.RData) and the novel Rstudio projects (.Rproj) for establishing dedicated working directories, workspace, history, and source documents.

Module II. Exploratory Data Analysis (EDA)

• During any statistical programming workflow, almost half of the time must be given to data exploration. However, not every exploration is a valid exercise. Here we will review the most common assumption in a classical statistical analysis like normality, heterogeneity and independence in the data.

Module III. Univariate Statistical Analysis (UniStat)

• In module III we will review the classic univariate approach for statistics like te linear models and their extensions. The most common analysis like ANOVA, ANCOVA or Regression analysis will be covered. For those common cases in ecology, where the linear models fail (e.g. non-negative data in count/abundance data) we will present some extensions covering the Generalized Linear Models (GLM) for count and presence/absence data and we will review some insights in Generalized Additive Models (GAM).

Module IV. Multivariate Statistical Analysis (MultiStat)

• Along with this fourth module, we will cover the analysis of multivariate data. We will review two different approaches to understand community (multivariate) data based on different ordination techniques. Thus, two approaches from unconstrained ordination, like Principal Component Analysis (PCA) and non-Metric Multidimensional Scaling (nMDS) will help us to reveal patterns along with our community data. Finally, we will use two approaches from constrained ordination like Redundancy Analysis (RDA) and Canonical Correspondence Analysis (CCA) to understand the role of some environmental (explanatory) variables over the community structure in our community Data.