Mining the student assessment data: Lessons drawn from a small scale case study

Abstract. In this paper we describe an educational data mining (EDM) case study based on the data collected during the online assessment of students who were able to immediately receive tailored and elaborated feedback (EF) after answering each of the questions in the test. Our main interest as domain experts (i.e. educators) is in studying (by employing any kind of analysis) how well the questions in the test and the corresponding EF were designed or tailored towards the individual needs of the students. The case study itself is aimed at showing that even with a modest size dataset and well-defined problems it is still rather hard to obtain meaningful and truly insightful results with a set of traditional data mining (DM) approaches and techniques including clustering, classification and association analysis. 1