Interplay of cognition and affect in undergraduate math students??? careers: insights from recursive partitioning

Data collected in entrance tests for undergraduate curricula in mathematics at the University of Turin are analysed using the recursive partition method, to obtain classification trees for different "response variables" describing academic achievement or drop-out. The input factors include both math abilities and several affective and motivational factors, the latter having being assessed using internationally validated questionnaires. We argue that classification trees can provide unexpected insight into the interplay of such factors for academic success or failure, specifically for math students.