An Examination of Affect and its Relationship with Learning among Students using SimStudent

The goal of this paper was to examine affect-related factors and its relationship with student learning while tutoring an agent called SimStudent. These affect-related factors are the negativity of student self-explanations, the incidence and persistence of student affective states. Secondary school students who were part of this study were asked to teach their SimStudents solve algebra equations and make them pass all the quizzes. Results revealed that students failed to learn which led us to investigate other factors that could have attributed to this failure. Although the non-negatively valenced self-explanations did not have significant relationships with the students’ learning gains, the self-explanations were helpful in terms of mathematical content and they generally exhibited positive attitudes when giving the self-explanations. Students also tended to perform better with higher levels of good confusion. Higher levels of boredom were associated with poorer learning. Boredom and confusion were the most persistent but both did not have significant relationships with student learning. Though the negative correlations of the negative self-explanations, incidence and persistence of boredom vis-à-vis learning were not significant, the findings imply that negativity is linked to students’ poor performance