The community of inquiry framework and learning analytics: A systematic review of previous research

• Learning analytics (LAK) has been widely used in Community of Inquiry (CoI) research and LAK • The integration of LAK into CoI research relies on text, log, and survey data • Most LAK-informed CoI research is predictive and/or descriptive • The CoI framework primarily informs research design and processes • LAK offers unique insights into the evolution of the CoI framework • Transparency in reporting and open science practices should be encouraged The purpose of the current systematic review is to provide a comprehensive overview of how the Community of Inquiry framework and learning analytics have been informing each other, thus providing suggestions for how to enhance future research and practice in online education. Overall results revealed that (a) research was primarily conducted in MOOCs and traditional online courses; (b) text and log data were the primary sources analyzed using various statistical, computational, and/or machine learning methods through various tools and software; and (c) descriptive and/or predictive analytics were the most common learning analytics methodology thus describing and/or predicting student and instructor outcomes including teaching, social and cognitive presence. Even though few studies have explicitly named it as the guiding framework, the Community of Inquiry framework has significantly influenced learning analytics research design and some studies have offered new theoretical insights. As for research quality or characteristics, fewer studies fully reported such important details as participant characteristics and data preprocessing procedures. All these findings led to the conclusion that the Community of Inquiry framework and learning analytics have been mutually beneficial so far, and similar future research needs to pay more attention to reporting quality thereby providing richer insights into both the theory and practice of online education.