Using causal models to bridge the divide between big data and educational theory
A new journal paper by Kirsty Kitto, Ben Hicks and Simon Buckingham Shum has just been published in the British Journal of Educational Technology:
Abstract: With the proliferation of data in educational settings it is becoming even more important to ensure analytical results are grounded in reality. Ideally we would like to embed our theoretical concepts from the learning sciences in our data products in order to facilitate trust and interpretability. But getting from a well described theory to a theory formal enough to be operationalised using methods from Artificial Intelligence and Learning Analytics can be challenging. A new paper, Using causal models to bridge the divide between big data and educational theory, from CIC, outlines how causal models may help bridge this gap between our scientific knowledge and the applications of big data. The paper also introduces the causal modelling apparatus and explores how it may offer new insights to the field of Learning Analytics.