ABSTRACT
In this chapter we present statistical tests which are suitable if the researcher wants to test some special specific null hypotheses about the treatments which are also known as contrasts. We illustrate which R packages contain suitable implementations and how the most important functions of them work. Such approaches often lead to a multiple testing problem which inflates the overall type I error rate if not treated appropriately. We discuss multiple approaches which allow to control the so-called family-wise error rate for various situations like all pairwise comparisons or comparison with a control. The focus is on the application of these methods in R.
