Identifying Characteristics of Reflection Triggers in Data Science Ethics Education

Ethics education in data science aims to teach aspiring data scientists a critical reflective data science practice. However, university courses must bridge the gap between theoretic knowledge of ethics and ethical practice. Towards this, our research aims to understand how we can promote a critical reflective practice through reflection. We, therefore, investigate how data science students start reflecting and what constitutes reflection-triggering contexts in data science education. For this, we introduce a reflective essay assignment and propose a reflection-sensitive inductive content analysis to analyze it. Our findings based on seven student reflective essays suggest that important reflection trigger characteristics in data science ethics education include students’ expectations, a new insight, motivators for reflection related to expectations, teaching formats, and emotions. Our reflection-sensitive analysis is suitable for explorative analysis and creates transparency about existing sensitizing concepts.