Detect students' academic emotions in classroom: Measurement, self-perception and manifested behaviors

The emotions students experience in the classroom are likely to influence their learning in school. For purposes of helping student improve their learning, this study seeks to further the development of the use of automated analysis of observed data to understand students' emotions and during their learning activities. This study looks at linking behavioral data about emotions with self-report data about emotions with the intent to determine if observed behaviors are able to predict student's emotions. The ICT-supported Learning Emotion Scale was used for 57 students to measure their learning emotion while learning in an ICT environment, meanwhile, classroom observations were conducted for 57 students for six weeks in 4 classrooms, paying close attention to student emotional behaviors and emotional changes. The results found that three key groups of behavioral indicators happened frequently. One of the key findings was that students who volunteered to answer questions in class were observed to have an observable “enjoy” emotion 84.79% of the time. The discussion highlighted that the bridge function of students' emotion engagement between classroom observation and questionnaire data.