Evaluating human-automation etiquette strategies to mitigate user frustration and improve learning in affect-aware tutoring

Abstract

Human-automation etiquette applies human-human etiquette conventions to human- computer interaction (HCI). The research described in this paper investigates how to mitigate user frustration and support student learning through changes in the style in which a computer tutor interacts with a learner. Frustration can significantly impact the quality of learning in tutoring. This study examined an approach to mitigate frustration through the use of different etiquette strategies to change the amount of imposition feedback placed on the learner. An experiment was conducted to explore how varying the interaction style of system feedback impacted aspects of the learning process. System feedback was varied through different etiquette strategies. Participants solved mathematics problems under different frustration conditions with feedback given in different etiquette styles. Changing etiquette strategies from one math problem to the next led to changes in motivation, confidence satisfaction, and performance. The most effective etiquette strategies changed depending on if the user was frustrated or not. This work aims to provide mechanisms to support the promotion of individualized learning in the context of high level math instruction by basing affect-aware adaptive tutoring system design on varying etiquette strategies.