Despite the well-known fact that data mining algorithms work best on large data sets, the study focused on student data available to HEIs which is limited and clearly falls into the category of a small data set. Results show that small student data sets in the specific data mining analysis did not limit the use of data mining tools. Moreover the Weka tool, that was used as a comparative analysis to MS Excel, showed that by using decision tree models a high prediction accuracy (especially with the REPTree model) is obtained (the accuracy was verified during the test phase).