Grounding Principles & Theories
Classlet’s design is informed by a blend of learning science, human-computer interaction, and AI pedagogy. Every task, interaction, and environment is shaped by foundational theories that promote engagement, agency, and authentic learning outcomes.
🧠 1. Embodied Cognition Theory
Embodied Cognition posits that cognitive processes are deeply rooted in the body's interactions with the world. As Wilson (2002) asserts, cognition is “not just situated but embodied,” meaning the way we perceive, remember, and reason is shaped by our sensorimotor experiences. Barsalou (2008) extends this, arguing that all abstract thinking is grounded in bodily states and simulations.
Research supports this: when learners manipulate physical or virtual objects, they form more durable conceptual links. Johnson-Glenberg et al. (2014) found that high embodiment in mixed-reality simulations led to superior learning outcomes compared to low-embodiment conditions. Makransky and Petersen (2021) further showed that immersive environments that allow learners to enact concepts increased cognitive engagement and retention.
Classlet philosophy: We integrate embodiment through grab, rotate, and move tasks — letting students “do” rather than just “view.” Especially in VR, learners manipulate virtual objects using either motion controllers or keyboard equivalents. This ensures even low-end devices benefit from bodily interaction. We believe action-based cognition accelerates understanding, particularly in spatial and ethical scenarios.
🧠 2. Self-Determination Theory (SDT)
Self-Determination Theory (SDT), developed by Deci and Ryan (1985), frames motivation as emerging from three innate psychological needs: autonomy (freedom to choose), competence (feeling effective), and relatedness (feeling socially connected). When these are satisfied, learners are more engaged and intrinsically motivated.
Ryan and Deci (2000) emphasize that autonomy-supportive environments lead to greater persistence and enjoyment. In digital learning, Chen and Jang (2010) and Makransky et al. (2020) have shown that VR environments which promote agency and competence support self-regulated learning and persistence even in challenging tasks.
Classlet philosophy: We design scenarios that give students meaningful choice — such as deciding which route to take, when to retry, or how to respond. GPT avatars provide emotionally intelligent feedback to strengthen relatedness. We minimize penalty for failure, offering non-punitive retries and scaffolding to reinforce competence. Our approach treats students not as passive recipients but as autonomous agents.
🧠 3. Technology Acceptance Model (TAM)
The Technology Acceptance Model (TAM) by Davis (1989) explains how users come to accept and use technology. Two main factors — Perceived Usefulness (PU) and Perceived Ease of Use (PEOU) — shape intention and behavior.
Educational research extends this by showing that PU and PEOU in VR settings are influenced by system design, clarity of task goals, and user feedback. Lopez-Ozieblo et al. (2025) identified that adaptive guidance and low learning friction were central to students’ willingness to use immersive learning tools. Similarly, Wong et al. (2024) showed that presence and usability jointly predicted TAM adoption.
Classlet philosophy: To boost perceived usefulness, Classlet offers curriculum-aligned scenarios and clear pedagogical outcomes. To enhance ease of use, we provide guided GPT walkthroughs, intuitive interactions (e.g., click or gaze to trigger), and low-cognitive-load interfaces. For mobile users, simplified gestures and audio cues reduce onboarding effort. Our system aims for effortless immersion.
🧠 4. Speech Act Theory and Prospect Theory
Speech Act Theory, from Austin (1962) and Searle (1969), suggests that language is not just descriptive but performative — it does things: commands, questions, promises, affirmations. This is especially relevant in AI-driven learning, where avatars give instructions or encouragement.
Prospect Theory (Kahneman & Tversky, 1979) explains that people perceive gains and losses asymmetrically: losses feel more intense than equivalent gains. In education, framing feedback or decisions as potential gains or risks can shape learner engagement (Wang & Wang, 2021). Stark et al. (2024) showed that VR learners gave more elaborate responses when prompts used loss-framed wording.
Classlet philosophy: Our GPT avatars use speech acts purposefully: directives to guide (“Try selecting…”), expressives to praise (“That was insightful!”), and commissives to motivate (“Let’s solve this together”). We alternate between gain- and loss-framed messaging to support decision-making and reflection. This strategy nudges learners toward higher cognitive elaboration.
🧠 5. Constructivism & Problem-Based Learning (PBL)
Constructivism, rooted in Piaget (1952) and Vygotsky (1978), views learning as the construction of meaning through interaction and experience. Learners build new knowledge by linking it to existing schemas, especially through social interaction and challenge.
Problem-Based Learning (PBL) applies constructivism through real-world, open-ended problems that learners must investigate, reason through, and reflect on. According to Choi et al. (2025), PBL in immersive learning is most effective when paired with structured reflection and meaningful choice. Wong et al. (2024) found that in branching VR simulations, learners showed greater depth of reasoning when required to explain choices to avatars.
Classlet philosophy: We use nonlinear task flow, contextual dilemmas, and reflection prompts to promote active construction. GPT avatars often challenge students to justify decisions or anticipate outcomes. Scenes are scaffolded to align with exploration → interaction → reflection loops. This mirrors Kolb’s experiential cycle and supports authentic reasoning across disciplines.
🧩 How These Shape the Experience
Embodied Cognition
3D grab/select, VR immersion
Self-Determination
Choice-based task flow, roleplay
TAM
Simplified UI, low onboarding load
Speech Act Theory
GPT agents mimic teachers, peers, guides
Problem-Based Learning
Multi-step challenges and simulation
Last updated