More common mistakes in the practical usage and interpretation of recursive partitioning approaches are the confusion of main effects and interactions (see, e.g., Berk 2006) as well as the application of biased variable selection criteria and a significance test for variable importance measures (see, e.g., Baca-Garcia et al. 2007) that has recently been shown to have extremely poor statistical properties.
