Students’ emotions and cognitive loads in chatbot-versus peer-supported reading

Abstract

Using a within-subject design, this study engaged 60 Chinese EFL university students in reading English academic passages either with GenAI chatbot support or with peer support. Employing questionnaires, we compared the levels of positive and negative emotions and intrinsic and extraneous cognitive loads reported by students in these two reading conditions. Additionally, semi-structured interviews were conducted with 16 students to delve into the factors affecting their emotions and cognitive loads in the two interactive reading conditions. The results of questionnaires show that students experienced a significantly higher level of positive activating and deactivating emotions such as hope and relief and a lower level of negative activating emotion such as anxiety, as well as significantly lower levels of intrinsic and extraneous cognitive load in chatbot-supported reading. The interviews further reveal that students attributed the observed advantages of reading with chatbot to its immediate support, high efficiency, convenience of tracking the discussion record, and the low-stress learning environment. In the meantime, they reported disadvantages in emotional communication and social dynamics when communicating with AI chatbot. Overall, this study has elucidated the mechanisms underlying differential effects of reading with GenAI chatbot support versus peer support on students’ emotions and cognitive loads.