Title: Analyzing big data on early literacy acquisition based on student interactions with digital reading supplement
Authors: Yawen Ma - Lancaster University (United Kingdom) [presenting]
Anastasia Ushakova - Lancaster University (United Kingdom)
Kate Cain - Lancaster University (United Kingdom)
Harrison Gamble - Amplify (United States)
Jennifer Zoski - Amplify (United States)
Abstract: The focus is on the development of an exploratory analysis framework to be used on a large and complex dataset that captures the interactions of school children with a digital reading support supplement. The aim is to understand what contributes to early literacy acquisition using big data. The data arrives comes from student interactions with a research-based game environment, Amplify, which is composed of a variety of interactive games designed to support various reading skills. There is evidence that performance on assessments of various reading skills (e.g., morphological awareness, word reading) are interdependent and reciprocal, therefore, making which makes modelling the predictors of reading comprehension for beginner readers challenging. To characterize children's performance in various games we will evaluate how longitudinal models (e.g., latent growth curve) could be used. We will then use the results of the models to define if there are emerging clusters of behaviors. We can then look at the interdependence of game performance within those clusters using a simple graphical model structure. We will provide the foundation not only for a novel framework that can be applied to reading games big data but also will help to answer pressing research questions within literacy research (e.g., how different skills interplay dynamically in children's learning).