Classification and regression trees (C&RT or CART) are a recursive partitioning method used to predict dependent variables that are either continuous (regression) or categorical (classification). The classic algorithm was developed by Breiman et al. (1984). Data are assigned into binary categories based on a set of logical if-then statements (the branches of the “tree”). Rules are selected to differentiate observations as much as possible based on the dependent variable. Once a rule is selected, the same process is applied to each node. Partitioning stops at a set point or when no more useful divisions can be made. C&RT is a useful method because the decisions are often intuitive, it is simple enough that new observations can be classified easily, and it makes no assumption that the underlying relationship between the dependent and independent variables is linear.