Faces Don’t Lie: Analysis of Children’s Facial expressions during Collaborative Coding

Abstract

Educational research has used the information extracted from facial expressions to explain learning performance in a variety of educational settings like collaborative learning. Leveraging this, we extracted the emotions of frustration, confusion and boredom from videos with children aged 13-16 years old while they were collaborating to create games using Scratch. After we computed the groups’ coding performance, based on the created artifacts, we divided them into high and low performance and compared them on the basis of individual emotions’ duration and the transitions among the emotions. The results show that the children from the high performing teams show more confusion and frustration and more often from confusion and frustration to delight and neutral. The low performing teams show more boredom and move to this emotion from any other. Based on the results, we suggest implications both for the instructors and students.