Dr. Jacolien Van Rij, Dr. Fabian Tomaschek
Recent statistical methods, such as generalized additive models (GAM; Wood, 2006) or functional data analysis (Ramsay & Silverman, 2002), allow for modeling nonlinear effects and interactions. The visualization of the effects is crucial for a correct interpretation of analyses with non-linear effects and non-linear interactions. Most of packages that implement statistical methods in R provide plotting functions, but in many situations these are restricted in use and not easily adapted to the user’s preferences. In EEG research, for example, the negative values are plotted upward by convention, but the plotting functions typically do not support this.
This hands on workshop offers an introduction in plotting statistical effects and interactions in R based on the model’s predictions. We will cover different types of plots, ranging from simple line plots to rotating 3D surfaces. A GAM model will be used as an example in this workshop. Knowledge about GAMs or complex regression models is not required, but a basic understanding of regression analysis is necessary. For this workshop, it is imperative to install the latest R version is required, as well as the latest versions of the following packages: mgcv, car, rgl, caTools.
Ramsay, James O., B.W. Silberman (2002): Applied functional data analysis – Methods and Case Studies. Springer-Verlag. New York, Berlin, Heidelberg.
Wood, Simon (2006): Generalized additive models – an introduction with R. Chapman & Hall/CRC. Boca Raton, London, New York.