Measuring Difficulty Of Introductory Programming Tasks

Quantification of the difficulty of problem solving tasks has many applications in the development of adaptive learning systems, e.g., task sequencing, student modeling, and insight for content authors. There are, however, many potential conceptualizations and measures of problem difficulty and the computation of difficulty measures is influenced by biases in data collection. In this work, we explore difficulty measures for introductory programming tasks. The results provide insight into non-trivial behavior of even simple difficulty measures.