The main goal of this framework is to provide the institution with a mechanism to capture a new stream of data from students that gives the system and instructors some insight into students’ affective state. In the future, this could enable each student’s learning to be automatically altered and aligned with their affective state to create a more effective learning experience.
The feature is designed to be lightweight to allow students to self-report their state at a very high level. Engagement with this feature is optional but they will be prompted at regular intervals to input and update their affective state.
A default framework which attempts to capture student mood is available. However, the framework is designed to be generic so that this can be adjusted or replaced to suit the requirements of each institution.
The default framework builds on the work of (Lang, 1980) and (Bradley & Lang, 1994), who used a discrete manikin-based 5-point scale to capture students’ self-reported position on each dimension in Pleasure-Arousal-Dominance model. We simplify the model from the three 5-points scales to a single scale represented by the four emoji faces shown below. This final model has been tested with data collected using Amazon’s Mechanical Turk, and has been shown to be capable of sufficiently capturing a wide range of possible emotions and affective states. 