Sample Report · Public Demo

Sample Thinking Trace Report

A Grade 3 student solving a 4 × 6 multiplication mission.

Level Grade 3
Accuracy 88%

What you're looking at

Inquiry AI records every hint, hesitation, and wrong answer during a Socratic mission, then turns that trace into a parent- and teacher-readable diagnosis. The report below is generated from a fixture trace — a real student session looks identical. Crucially, the analysis runs entirely on pre-authored content with zero runtime LLM calls: it is repeatable, transparent, and child-safe.

Offline-firstHesitation trackingMisconception mapCCSS aligned

How the diagnostic reads a session

A thinking trace is more than a score.

This public demo gives Google and families a crawlable example of the Inquiry AI diagnostic layer: what the child tried, which misconception appeared, and which next practice page should follow.

Hesitation

The student paused before choosing a model.

A long pause before the first answer is treated as a reasoning signal, not as failure. The report can separate slow setup from an actual misconception.

Misconception

The first wrong answer confused groups and items per group.

For a 4 x 6 array, answering 10 suggests adding the two visible quantities instead of reading rows and columns multiplicatively.

Recovery

The student corrected the strategy after a targeted hint.

Recovery time matters because it shows whether a Socratic prompt helped the child rebuild the idea without an answer reveal.

Thinking trace diagnostic demo

Who this page helps, and where to go next.

A public sample of a math diagnostic report for parents and teachers who want to see hesitation, misconception, recovery, and standards alignment instead of an answer-only score.

Best for

  • Parents comparing math apps and AI tutors who need proof of how a child is thinking.
  • Teachers looking for a readable diagnostic layer before assigning more practice.
  • Tutors and homeschool adults who need to separate careless errors from concept gaps.

Problems solved

  • Most practice tools show right or wrong but hide the reasoning path that produced the answer.
  • A child can recover after a hint, but the adult cannot see which prompt unlocked the work.
  • Runtime AI explanations can be inconsistent, so families need a transparent sample of pre-authored Socratic diagnosis.

No cognitive traces recorded yet. Complete missions to see analysis.

CCSS Standards Alignment

3.OA.A.1

Interpret products of whole numbers.

3.OA.A.2

Interpret whole-number quotients.

3.OA.A.3

Solve multiplication & division word problems.

3.NBT.A.3

Multiply one-digit numbers by multiples of 10.

Want a report like this for your child?

Sign in to start a real Socratic mission. Your child's traces stay private — only you see the diagnosis.