From raw to ready-made data. A hands-on manual for pre-processing Learning Management System log data for learning analytics

A lot of data is being generated online as a by-product of human behavior. These data can give interesting insights for social scientists. However, since these data are often stored in large ›raw‹ log data tables, collecting and analyzing the data is not straightforward for those inexperienced in handling such files. In the current chapter, we outline a procedure for pre-processing raw data into an analyzable format. As our example, we use data from students in a learning management system. Data from are often used to gain insight in students’ online learning behavior (»learning analytics«). The chapter provides a hands-on manual for pre-processing data from Moodle, using the opensource software, scripts are provided for every pre-processing step, and we highlight decisions that have to be made during the process. The whole process is designed in such a way that researchers with little background in advanced data manipulation techniques are able to transform their own data with relatively little own input. As a proof of concept, the chapter ends by using the preprocessed data for a common analysis in the field of learning analytics: predicting student performance.