FirstBite Framework V2.0

A learner-centered, research-driven iterative model for behavior-change cooking apps

Model Overview

FirstBite Framework

Phases & Activities

🌱
Phase 1: Discovery & Grounding (1.5 weeks)
  • Conduct deep learner archetype workshops, not just surveys (co-design with students!)
  • Shadow students during grocery shopping, cooking attempts
  • Contextual inquiry: Map routines, pain points, tech comfort
  • Review literature on cooking confidence, health literacy, time scarcity in youth
What's better: More real-world immersion + evidence-backed foundations
🎯
Phase 2: Learning & Behavior Goal Mapping (1 week)
  • Use backward design: start from long-term behavior goals (e.g. autonomous meal planning)
  • Develop dual pathways: one for skill acquisition, one for habit formation
  • Define performance benchmarks + reflective goals (self-efficacy, identity)
What's better: Aligns UX with motivational and behavioral learning theory
🧠
Phase 3: Cognitive Flow & Strategy Mapping (1 week)
  • Break cooking tasks into cognitive chunks using worked examples
  • Align content with Cognitive Load Theory: simple-to-complex scaffolding
  • Define AI intervention logic: when to guide, when to step back
What's better: Builds smart adaptive strategies before designing UI
🧪
Phase 4: Rapid Multimodal Prototyping (2 weeks)
  • Develop low- to mid-fidelity prototypes across modalities (chatbot, video, tactile prompts)
  • Test versions with different delivery formats (flashcards, timers, VR kitchens!)
  • Pilot one cooking journey with scripted learner paths
What's better: Prototyping is not app-first, but experience-first
🔁
Phase 5: Agile Test Loops (3 weeks)
  • Deploy Wizard-of-Oz simulations before AI integration
  • Weekly testing sprints with 5–7 learners each
  • Assess engagement via biometric proxies (optional): pace, hesitation, vocal tone
What's better: High-fidelity learning feedback before full app development
📊
Phase 6: Holistic Evaluation & Narrative Impact (2 weeks)
  • Mix Kirkpatrick + Transformative Learning Theory reflection tools
  • Capture video journals, photo food diaries, and AI-facilitated insight prompts
  • Deliver report to stakeholders with visual storytelling + data dashboard
What's better: Evaluation includes storytelling, motivation, and decision data

Reflection

In this ideal model, we go far beyond what we did in our current DPL project. Instead of designing fast to ship, we design deeply to understand. We embed theory before tools, involve learners as co-creators from Day 1, and emphasize realistic behavior change over polished UI.