Scoring Guide (applies to ALL components)
5 — Highly Authentic
Fully realistic and grounded in real student behavior and learning dynamics.
4 — Mostly Realistic
Believable with minor gaps or underdeveloped learning nuance.
3 — Plausible but Generic
Possible, but flattened; lacks specificity or human learning “mess.”
2 — Weak / Artificial
Noticeably constructed, overly polished, or misaligned with how students behave.
1 — Unrealistic / Broken
• Persona — Is this a real student?
Do the age, background, learning level, and behavior feel realistic and internally consistent?
• Intent — Is this how students express what they want?
Does the request sound human and imperfect (confused, anxious, overconfident, vague), even if the student is wrong about what they actually need?
• Scenario — Is this a real learning situation?
Does the context reflect how learning actually happens?
Consider:
• time pressure
• gaps in foundations
• exam or performance stress
• motivation or disengagement
• Hidden Context — Is something unsaid but important for learning?
Is there a realistic hidden factor affecting how teaching should happen?
Examples:
• misconceptions
• fragile confidence
• surface memorization
• fear of failure
• Initial Prompt — Is this how students really ask for help?
Does it sound natural, informal, or emotionally loaded — like something a student would actually say?
2️⃣ Trap Authenticity (Critical)
Using the same 1–5 Scoring Guide, evaluate whether the scenario creates a real pedagogical trap for the AI.
Does it embed non-obvious learning constraints that could mislead the AI unless it actively diagnoses the student (cognition, emotion, confidence, timing)?*
3️⃣ Expert Rationale (Required)
After scoring, write a clear expert justification:
• “Persona scored X because…”
• “Intent scored X because…”
• “Scenario scored X because…”
• “Hidden Context scored X because…”
• “Trap Authenticity scored X because…”
Good Luck!
