Physiological Computing Education: Towards Teaching Using Physiological Signals
Physiological computing is a novel technology revolutionizing fields outside medical disciplines. Recently, there are multiple implementations of physiological data in educational contexts (e.g., user state monitoring and adaptive experiences). The literature shows increased interest in creating systems for novice programmers to learn the development of closed-loop neurofeedback applications. These types of signals can engage learners, increasing their interest in the technology. However, the back-end processes necessary to interpret physiological signals regularly require Machine Learning classification algorithms, which presents an opportunity to include ML-based concepts during educational activities. Machine Learning (ML) education is becoming important and necessary in computing education. Nowadays, everyone experiences this technology through everyday software. Social media, streaming services, online stores, and home devices (e.g., Google Home and children's smart toys) are accessible by anyone. These trends motivate current research to create explainable educational systems for ML to increase the number of people that can contribute to the field. This presents an opportunity to create physiological-based systems that leverage the ML-based architecture to expose and teach users about the technologies. This dissertation presents a set of studies investigating the impact of gamification and physiological signals in the learner's interest, feelings, confidence, and knowledge. The first study explores the effects of a BCI educational platform on the self-efficacy of novice users. The second study presents the first iteration of PhysioML (a gamified educational application using physiological signals) and its evaluation through a within-subjects design. The third study expands upon PhysioML to evaluate its effects on learning, self-efficacy, stress, and system usability. Results suggest that this system can be an effective tool for novice ML and physiological computing developers. Participants showed positive learning outcomes and a better understanding of real-world solutions connected to the topics. The objective is to provide a foundational piece of literature and a novel educational platform that motivates the exploration of ML-powered physiological education platforms created to lower the barriers to entry in this field.