Affect and Behavior in Digital Classrooms (PhD project)
How can we employ process data to infer the affective states and behavior of children interacting with digital learning tools, and what is the impact of affect and behavior on learning outcomes?
About the project
This PhD project builds on “The Vocabulary Learning Challenge” (VLC). In this project, we developed a morphology-based app to support vocabulary learning in second-graders. More information on the VLC project can be found here.
The first part of the PhD project builds on the randomized controlled trial study conducted in the VLC project. As the vocabulary learning app is based on morphology, morphological knowledge is a key concept when measuring outcomes. Since morphological knowledge is a very complex construct, we are conducting a study that investigates the how we conceptualize, measure and analyze this ability. This study looks into the dimensionality of morphological knowledge; whether it is best viewed as a single independent ability or several interrelated abilities.
In the second part, we aim to develop automatic detectors of children’s affect and behavior. To achieve this goal, we will use machine learning techniques along with field observations to classify affect and behavior based on patters in the process data from the app. Furthermore, we wish to investigate what impact different behaviors and affective states have on children’s learning outcomes from the app.
The project started in March 2019 and is expected to finish in March 2023. Main supervisor in the project is Bjõrn Andersson (CEMO) and co-supervisor is Janne von Koss Torkildsen (ISP).