Research foundation

Misconception research has decades of evidence. We built on it.

The idea that students form stable, internally consistent alternative mental models — not random errors — is one of the most replicated findings in education research. It was documented by Chi and colleagues in 1981, extended to mathematics by Confrey in 1990, and applied to formative assessment design by Black and Wiliam in 1998. Brainpathio's engine is the applied version of what that research recommends.

The three pillars

The research our taxonomy is built on

Pillar 1

Misconceptions as stable conceptual structures

The foundational insight: student errors are not random. Chi, Feltovich and Glaser (1981) demonstrated that novice learners organize knowledge into stable but incorrect conceptual frameworks — frameworks that are internally consistent from the student's perspective. Confrey (1990) extended this to mathematics, showing that fraction errors like "1/2 + 1/3 = 2/5" reflect a coherent theory about how fractions work, not a random mistake. Smith, diSessa and Roschelle (1993) established that these frameworks are resistant to change through direct instruction alone — they require targeted conceptual disruption.

How Brainpathio applies this: We don't treat wrong answers as "incorrect attempts at the right answer." We treat them as evidence of a specific alternative model, and our hypothesis engine tries to identify which model best explains the error pattern.

Chi, M., Feltovich, P., & Glaser, R. (1981). Journal of Educational Psychology · Confrey, J. (1990). Journal for Research in Mathematics Education · Smith, J., diSessa, A., & Roschelle, J. (1993). Journal of the Learning Sciences

Pillar 2

Algebra-specific misconception taxonomy

Kieran (1992) systematically catalogued algebra misconceptions in early algebra learners, distinguishing between structural (conceptual) and procedural errors. The "variable as label" misconception — where students treat algebraic letters as abbreviated object names rather than unknown quantities — was identified by Küchemann (1981) as the single most prevalent block to algebra progression in Grades 6-8. Sfard (1991) introduced the operational/structural duality framework: students may understand mathematical objects operationally (as processes) but fail to understand them structurally (as entities), which explains why a student can "do" a procedure but not explain or transfer it.

How Brainpathio applies this: Our Algebra category draws directly from Kieran's taxonomy. The "variable as label" misconception is the highest-priority category in our Grade 6-8 engine, as it blocks more downstream mathematics than any other single misconception.

Kieran, C. (1992). In D. Grouws (Ed.), Handbook of Research on Mathematics Teaching and Learning · Küchemann, D. (1981). In K. Hart (Ed.), Children's Understanding of Mathematics · Sfard, A. (1991). Educational Studies in Mathematics

Pillar 3

Diagnostic assessment and formative feedback loops

Black and Wiliam's (1998) landmark synthesis of 250+ studies established that formative assessment — when used to identify specific learning gaps and adjust instruction accordingly — is one of the highest-effect-size interventions available to classroom teachers. Heritage (2010) identified the critical gap: most formative assessment tools tell teachers who is struggling but not why, making it impossible to close the feedback loop effectively. The NRC's "How People Learn" (2000, updated 2018) frames effective instruction around activating and correcting prior knowledge — the misconception — rather than simply layering new information on top.

How Brainpathio applies this: Our alert feed is designed specifically to close the gap Heritage identified. We tell teachers not just that students are struggling, but which specific prior knowledge is blocking them — making formative assessment actionable.

Black, P. & Wiliam, D. (1998). Phi Delta Kappan · Heritage, M. (2010). Formative Assessment: Making It Happen in the Classroom. Corwin · National Research Council (2000). How People Learn. National Academies Press

Our own data · Preliminary

From our 2025-2026 pilot across 3 Portland-metro schools

Internal analysis. Preliminary findings. Sample sizes are small. These are signals, not statistically significant claims. We share them because transparency about what we've seen — and our uncertainty about it — is part of our epistemology.

72%
Of wrong answers attributable to a documented misconception category
In our pilot data, 72% of incorrect student responses matched one of our 42 documented misconception patterns — suggesting the remainder are genuine random errors or procedural slips.
3.4×
Faster misconception resolution with targeted vs. general reteach
When teachers used Brainpathio alerts to run targeted interventions, students cleared the misconception 3.4× faster than with standard whole-class reteach. Small n — interpret with caution.

Public taxonomy excerpt

12 documented misconception types — with research citations

A partial excerpt from the full taxonomy (42 types). Full taxonomy available to pilot participants.

DomainMisconceptionBrief descriptionResearch citation
AlgebraVariable as labelStudent treats x as an abbreviation for an object (x=apples) rather than an unknown quantityKüchemann, 1981
AlgebraInverse operation confusionDoes not recognize that solving requires applying the inverse operation to both sidesKieran, 1992
AlgebraEquality as operatorInterprets = as "compute answer" rather than "balance relationship"Alibali et al., 1999
AlgebraSign confusion in expressionsTreats -x as necessarily negative, rather than as the opposite of xVlassis, 2004
FractionsWhole-number interferenceApplies whole-number rules to fraction operations (e.g., larger denominator = larger fraction)Stafylidou & Vosniadou, 2004
FractionsPart-whole confusionConflates part-whole and ratio interpretations of fractionsConfrey, 1990
FractionsIndependent numerator/denominatorAdds or subtracts numerators and denominators independently across unlike fractionsKerslake, 1986
FractionsFraction as two numbersTreats a/b as "a and b" rather than a single quantityNi & Zhou, 2005
RatiosAdditive ratio reasoningUses additive comparison in a multiplicative context — compares differences rather than ratiosHart, 1981
RatiosUnit rate confusionCannot distinguish between rate and the unit rate (e.g., 60 miles per 2 hours vs 30 mph)Thompson, 1994
Physical ScienceForce = motionBelieves that a moving object has a force "in it" keeping it moving — motion requires continuous forceClement, 1982
Physical ScienceHeat as substanceTreats heat as a material thing that flows or can be "trapped" — rather than energy transferLinn & Songer, 1991

Request our full methodology document.

Pilot applicants receive the complete taxonomy (42 types), research citation map, and pilot methodology document. No obligation.