Applying Data Analysis to Identify Early Indicators for Potential Risk of Dropout in CS Students

Dropout rates in STEM subjects are consistently high almost all over the world. To remedy the situation, many universities are taking measures that aim at reducing dropout rates. For these measures to be most effective, tools are required to predictively identify those students that are at risk of dropping out, to identify the individual causes of their struggles, to then select and apply appropriate interventions and finally to assess the effects that these interventions achieve.In this study, we focus on the first of these steps, i.e., identifying early indicators for potential risk of student dropout in our CS related degree programs. To achieve this, we analyse the data from our student administration system, focusing on our students’ performance in their first semester of studies. More precisely, we relate dropout to success rates and investigate into different behavioural patterns that lead to student dropout. Furthermore, we identify several factors that increase the risk of dropping out, and document some first early indicators for student dropout.