Affective learning analysis of children in a card sorting game

Abstract

The purpose of this paper is to provide an affective learning analysis on children while they were playing a card sorting game. The electroencephalogram (EEG) signals of 8 preschoolers aged between 4 to 6 years were collected (a) while they were playing a card sorting game; and (b) observing affective faces. The features from EEG signals were extracted using Kernel Density Estimation (KDE). The Multi-Layer Perceptron (MLP) was used to classify and generate the affective maps of the EEG signals while the children were playing the game. The initial results show that the children’s affective states are unique and there might be different affects that drive a child’s performance. This analysis shows the potential of using the affective learning analysis approach in assessing educational tools such as computer games.