MAD Learn: An Evidence-based Affordance Framework to Assessing Learning Apps

Abstract

Existing recommendations about how to select or design mobile applications (apps) for learning have been heavily relied on customer and teacher reviews, designer descriptions, and educational theories. There is a lack of evaluation frameworks that are informed by research evidence of how different children interact and use apps. The first version of an evidence-based framework, coined as MAD learn, is presented detailing affordances that hinder or help children’s learning, as emerged from relevant studies. To encourage further studies in the field, not only by researchers but also designers and practitioners, a methodological approach to iteratively assess the affordances of mobile apps is also introduced. This is based on (a) visualising the learning design and learning components of a given app and (b) analysing the screen and audio recordings of children’s interactions with apps. The proposed approach has been tested with 17 children 5–6 years old who interacted with a maths app. The analysis captured patterns of actual usage, including time spent on different activities, completion rates, communication instances, and number and type of mistakes. Insights revealed that certain design affordances, including instructions, feedback, and help-on-demand, were differently perceived by children, in some cases helping to learn while, in others, hindering it.