Delgado-Gomez et al. (2016) have provided an interesting example of the comparison of computerized adaptive testing (CAT) based on unidimensional item response theory (IRT) and a machine learning approach based on decision trees (DT) in the prediction of people who have a history of suicide attempts. ... In the following we attempt to better articulate the differences between the two statistical approaches so that practitioners and consumers of these methods will have a better understanding of their utilities.