How the engine works

Every wrong answer has a reason. We find it.

Brainpathio's misconception engine builds a hypothesis about why a student answered incorrectly — then tests it with the next question. Each student follows a different path based on their specific conceptual break.

See it with your curriculum

Misconception taxonomy

40+ documented misconception categories per grade band

Each category is drawn from peer-reviewed research and classroom observation. Not invented — documented. Four domains are currently active; Geometry and Life Science are in development. Brainpathio does not cover all of K-12 STEM — we cover the domains where misconception research is deepest and where our taxonomy has been validated against pilot data.

Algebra Misconceptions

Variable as label, inverse operation confusion, equality as operator, sign handling errors, distributive property misapplication. 12 types documented, grades 6-12.

Fraction Misconceptions

Whole-number interference, part-whole confusion, fraction as two separate numbers, unlike denominator addition. 10 types, grades 5-8.

Ratios & Proportions

Additive reasoning applied to multiplicative contexts, ratio as subtraction, unit rate confusion. 9 types, grades 6-9.

Physical Science

Force equals motion misconception, heat as substance, speed-velocity conflation, gravity as a pull-only force. 11 types, grades 6-12.

Engine routing logic

How a wrong answer becomes a targeted diagnosis

1

Student response

Wrong answer detected

The engine records the specific wrong answer, not just "incorrect." The answer choice itself carries diagnostic information — different wrong answers map to different misconception hypotheses.

2

Hypothesis activation

2-4 misconception hypotheses activated

Based on the question type and wrong answer, the engine activates the most likely misconception hypotheses from the taxonomy. A prior probability is assigned to each based on grade-level frequency data.

3

Diagnostic question

Next question tests the top hypothesis

The next problem is selected to maximally distinguish between the top two hypotheses. The student's answer updates the probability distribution using Bayesian inference.

4

Profile update + alert

Profile updated; teacher alert if threshold reached

If a misconception reaches 70%+ confidence, the student profile is flagged and the teacher receives an alert with the specific misconception name and a suggested intervention.

Worked example

Full walk-through: 7th grade algebra problem

One problem. Three possible root causes. Three completely different teacher interventions. Here's the full sequence.

Algebra · Grade 7

Solve for x:   2x + 3 = 11

Student selects: x = 7   (Correct: x = 4)

Route A: Inverse Operation Confusion
Student subtracted 3 from 11 (got 8) then subtracted again — didn't recognize division as the next step.
Diagnostic Q: "If 3x = 12, what is x?" — tests whether student can isolate a variable using division
Teacher intervention: Inverse operations concept map activity (~10 min)
Route B: Variable as Label
Student treated x as a placeholder (like "2 apples + 3 apples") and counted up to 7, treating x as having a fixed meaning from context.
Diagnostic Q: "In the equation n + 5 = 9, can n be any number?" — tests variable concept
Teacher intervention: Variables as unknowns conceptual discussion (~20 min)
Route C: Arithmetic Slip
Student correctly isolated to 2x = 8 but computed 8÷2 = 7 (arithmetic error). Algebraic understanding is intact.
Diagnostic Q: Simple arithmetic drill — 8÷2, 10÷5 — no algebra. Confirms computation, not concept.
Teacher intervention: None for algebra. Optional: mental math practice.

See it with your curriculum — not ours.

Apply for an 8-week pilot and watch the engine work with your students' actual error patterns.

Apply for Pilot Read the research