Product
A pass here means something.
Completion tells you a video finished playing. StudAI Loop gates every chapter behind four kinds of evidence — and tells you the thresholds up front.
The four dimensions
Every chapter's mastery gate measures all four. Each has its own published threshold, and all four must pass — there is no averaging away a weakness.
Recall
threshold 0.7Can you remember the facts, syntax, and definitions?
Transfer
threshold 0.6Can you apply the concept to a situation you have not seen?
Error Recognition
threshold 0.5Can you spot what is wrong in flawed work?
Explanation
threshold 0.5Can you explain it clearly enough to teach someone else?
Failure is a feature, not a shame
When a dimension fails, Loop diagnoses why and re-teaches those exact concepts with a different pedagogy — then re-tests only what failed.
- 1. DiagnosisWhich concepts, which dimensions, and the likely misconception behind each miss.
- 2. Targeted remediationMicro-lessons in a different format than the original teaching, followed by a focused mini-quiz across all four dimensions.
- 3. Bounded honestyA maximum of 3 cycles. After that, a human-supported assist lane — never a silent pass.
Evidence you can stand behind
Mastery results are stored as immutable, transactional evidence records — the same records your final credential exposes for public verification. Practice content generated during an AI outage is clearly labeled and can never create mastery evidence.
- Why did I fail with a high overall score?
- Because mastery is not an average. Each of the four dimensions has its own threshold, and all four must pass. Being excellent at recall cannot compensate for being unable to apply or explain a concept.
- Can I retake a quiz?
- Yes. Retries are supported, and identical resubmissions are idempotent — retrying the same attempt never duplicates or corrupts your evidence. Failed dimensions route you to remediation rather than blind repetition.
- Is the AI grading fair?
- Objective questions are graded deterministically — no AI involved. Open-ended answers are graded against a rubric with schema-validated output, evidence minimums, and confidence intervals that prevent one lucky or unlucky answer from flipping your result. Certification results carry an appeal path reviewed by a human.
- What happens after three failed remediation cycles?
- You enter the assist lane: a support case with a visible ID and status, alternate resources, and a supported retry when you are ready. Your record never claims mastery you have not demonstrated.