Learning enhancing emotions predict student retention : Multilevel emotions of Finnish university physics students in and outside learning situations

Research on student retention in higher education (HE) physics could benefit by studying emotions in the context of engagement and learning. However, popular retention theories include only a narrow selection of emotions, creating a need to look elsewhere. In this study, we borrow the lens of an affective engagement model, the framework of an optimal learning moment, which has rarely been used in HE research so far. It defines situational engagement and three categories of learning enhancing, detracting, or accelerating emotions via twelve singular situational emotions. These twelve emotions in learning and other situations form intensive longitudinal data collected from twenty students using the experience sampling method (ESM) during their first two months of studying physics in a Finnish university. A two -level hierarchical dataset consisting of ESM measures (N1 = 440) and student records (N2 = 20), with gender as a background factor, are analyzed in two steps: first with hierarchical linear modeling, followed by multinomial logistic regression, giving results on both levels of the hierarchy, which is quite uncommon still. The results show how situational engagement and learning situations are separately connected to situational emotions and, further, how especially the learning enhancing emotions connect to success in courses (passing, grades) and first year student retention, surpassing the effect of course success.Peer reviewe