Making sense of learning analytics with a configurational approach

Abstract

This paper is an attempt to provide the basic guidelines on how to implement configurational analysis in the context of learning analytics. In detail, we offer a step by step approach on the fuzzy set qualitative comparative analysis (fsQCA). Learning analytics gain increased popularity, however studies use traditional symmetric statistical methods to analyze them. Building on the theory of complexity and configuration theory we suggest on using fsQCA in order to gain a deeper understanding of the data, which may lead to understanding different learning phenomena as well as to the creation of new theories. We further describe the steps on how to perform a contrarian case analysis, which will help in identifying asymmetric relations among the data. Finally, testing for predictive validity with fsQCA is explained. Many of the steps described here may be implemented in various contexts, however we tried to provide examples and instructions for learning analytics oriented research.