The emotions experienced by primary school students have both positive and negative effects on learning processes. Thus, to better understand learning processes, research should consider emotions during class. Standard survey-based methods, such as self-reports, are limited in terms of capturing the detailed trajectories of primary school children's emotions, as their abilities of self-reporting are developing and still limited. Emotions can also be tracked by capturing emotional responses as they occur e.g. from physiological reaction measured with sensor wristbands. This technology generates an emotional responses typology based on continuously captured physiological data, such as skin conductivity and skin temperature. However, such measurement methods need to be validated before being used. The present study thus attempted to validate this instrument with primary school students. We used the BM Sensor Wristband technology, as its emotional response typology is based on the categorical emotion and homeostasis approach. In our research, we focus on the emotional responses that can be distinguished by the BM Typology and that can influence learning processes. These emotional responses are: "joy", "curiosity", "attention", "fear", "anger" and "passivity". Therefore, we induced emotional responses in primary school children through specifically developed audio-visual stimuli. Using logistic mixed effects modelling, we investigated the occurrence of opposing reactions. We observed that primary school children's reactions to audio-visual stimuli could be differentiated. We conclude that primary school children's emotional responses, such as "joy", "curiosity", "attention", "fear", "anger" and "passivity", can be accurately measured by evaluating physiological data.