Capturing Learning over Time for Supporting Pedagogical Decision Making

This chapter aims for a methodological contribution to online learning research and to the practical use of temporal information for pedagogical decision making. The authors address two interconnected concerns: how to describe the temporal features of teaching/learning activities and how to capture learning activities across learning applications and time. The main argument is that the analysis of temporal processes based on student data that can be automatically captured (in log files and through other means) will benefit from an explicit modeling of the teaching process, because in this way, some of the problems associated with a purely inductive approach to process and sequence mining can be overcome. In terms of advancing the state of the art, the authors suggest an approach that is grounded in meta-model architectures for process modeling and demonstrate its advantages with respect to tracking and monitoring students’ learning activities across learning applications. After providing some background on long-term learning, the chapter describes the conceptual as well as several of the implementation solutions developed in the EC-funded NEXT-TELL project.