Log generation for coding behavior analysis: For focusing on how kids are coding not what they are coding

Abstract

Block programming lowers the barrier for programming learners and it is used in many software education program. Based on our observation, we realized that there are differences in way of learning and time of finishing goals even in under same instructors. To know the cause of this difference we propose a logging function to see the coding behavior of programmers. In this work we have developed library for generating log of developer's behavior in the process of block programming and defined required common items in creating block log process. In addition, we present the coding characteristics from the log, available information for deriving coding characteristics and detail criteria for deriving each characteristic. The contribution of this work is in development of framework generating logs of block programming process. This work will contribute to understand the programming learners' behaviors and enable instructors to design the learning courses properly.