White Paper · Lacefield Pedagogical Framework · v2.0
Teachers & educational researchers Students & cognitive psychologistsA taxonomy of six instructional error types — from teacher subject-matter error to penalizing valid novel reasoning — that share a common damage mechanism: disrupting the student's trust in their own reasoning process. These errors are widespread, underreported, and not confined to early education. They occur at every level of instruction, in every subject, including at the highest levels of academic teaching.
This paper documents a taxonomy of six instructional error types that damage student confidence and mathematical development through a common mechanism: disrupting the student's trust in their own reasoning process. The taxonomy begins with teacher subject-matter error — the most basic failure mode, in which the teacher's own incorrect understanding propagates to students — and proceeds through curriculum-level inefficiency, three distinct feedback evaluation errors, and the inappropriate use of public questioning. The paper argues that these errors are not failures of poor or undertrained teachers alone. They occur at every level of education, including at the collegiate and graduate levels, wherever teacher subject-matter knowledge is insufficient relative to what they are teaching, wherever time pressure overrides careful evaluation, and wherever institutional culture tolerates practices that prioritize teacher convenience over student accuracy. The most severely underrecognized error in the taxonomy — penalizing a valid novel method — is addressed in depth, because it is simultaneously the most damaging and the most inverted in its reward implications: the behavior that should receive the strongest positive recognition in mathematics education is systematically treated as an error. The Hattie and Timperley (2007) feedback taxonomy provides the primary conceptual framework; Ball, Thames, and Phelps (2008) and Hill, Rowan, and Ball (2005) provide the evidentiary basis for the subject-matter knowledge requirement.
The common framing of incorrect correction as a problem of undertrained or low-level teachers is empirically inaccurate and practically harmful, because it prevents the mechanisms from being recognized in settings where they are causing damage. These errors do not require a poorly trained teacher. They require only a teacher whose subject-matter knowledge is insufficient relative to the depth of what they are teaching — which occurs at every level of the educational system — or a teacher who evaluates student work without examining the reasoning behind it — which is common at every level due to time pressure and class size — or an instructor who uses public questioning as a motivational tool — which is documented across primary, secondary, and higher education.
A graduate teaching assistant who completed the course two semesters ago is teaching content at the outer boundary of their current fluency. A lecturer covering twenty-five years of material will occasionally error on problems not recently worked. A professor managing a large class under time pressure will frequently evaluate student work at the answer level rather than the reasoning level. None of these require incompetence. All of them produce the damage documented in this taxonomy.
The taxonomy is documented primarily in mathematics because mathematics offers the most precise operationalization of its mechanisms — when a teacher gives a wrong answer, it can be proven wrong; when a student derives a valid novel method, its validity can be verified. The same mechanisms operate in any subject where teacher subject-matter knowledge has a depth requirement, where student reasoning can be examined, and where the difference between right-answer-wrong-reasoning and wrong-answer-right-reasoning is educationally significant. Which is most subjects.
"These errors are not confined to early childhood education or GED tutoring. Surprisingly high-level professors do petty things that discourage students and confuse them. The mechanisms are widespread, underreported, and under-corrected across all levels of instruction."
The depth of teacher subject-matter knowledge is not a peripheral variable in instructional quality — it is the prerequisite on which the avoidance of most of the errors in this taxonomy depends. A teacher who does not understand the subject deeply enough cannot reliably detect their own errors, cannot recognize valid novel methods from students, cannot distinguish a conceptual error from an execution error, and cannot present derivable content as derived rather than memorized.
Hill, Rowan, and Ball (2005) established this empirically in a study of first- and third-grade mathematics teachers. Teachers' mathematical knowledge for teaching was a significant predictor of student achievement gains at both grade levels — outperforming teacher background variables and time spent on mathematics daily as predictors. The key measure was not general mathematics knowledge but what Ball, Thames, and Phelps (2008) distinguish as specialized content knowledge: the mathematical knowledge unique to the work of teaching, which includes the ability to analyze the appropriateness of alternative methods, identify the origins of student misconceptions, and explain why a procedure works rather than simply that it works. This is distinct from common content knowledge — the mathematics a teacher shares with anyone who knows the subject — and it is the specific knowledge type most relevant to avoiding the errors in this taxonomy.
Ball et al. (2008) further distinguish horizon content knowledge — awareness of how mathematical topics are related across the span of the curriculum — as a separate domain. A teacher without horizon content knowledge cannot reliably identify when a student's alternative method, while unfamiliar, is mathematically valid. They have no framework for evaluating methods they have not seen before. This is the specific knowledge gap that makes mechanism 5 — penalizing valid novel methods — inevitable in teachers without sufficient subject-matter depth.
The implication for teacher preparation is direct and under-implemented. Early childhood and elementary education teachers in most credentialing programs take mathematics-for-the-liberal-arts courses that cover content at or slightly above the level they will teach, without developing the specialized or horizon content knowledge that teaching at that level actually requires. A middle school mathematics teacher who has taken two semesters of calculus and no abstract algebra does not have the depth required to recognize all valid alternative approaches to the content they are teaching. The issue is not the credential level — it is the specific content and the depth at which it is taught. A teacher needs to understand mathematics substantially beyond the level they are teaching to avoid the errors that insufficient depth produces.
The six mechanisms are ordered chronologically as they arise in instruction — from errors that originate in the teacher's own understanding, to errors in how student work is evaluated, to errors in how students are called upon publicly. All six share the common damage pathway documented in section 4.
The teacher works through an example or reviews student work and produces or endorses an incorrect answer. The error is presented with the authority of the teacher's role, which means students update their understanding accordingly. Students who had the correct answer now believe they were wrong. Students who had the wrong answer receive false confirmation. The damage is not confined to the specific problem — it propagates forward as a wrong schema into every subsequent problem that depends on that concept.
This mechanism is the most basic failure mode and the one most directly addressed by the subject-matter knowledge requirement. A teacher with sufficient depth in the subject makes fewer factual errors and, crucially, is more likely to catch their own errors when they occur. A teacher operating at the outer boundary of their subject-matter knowledge cannot reliably detect their own mistakes — the same knowledge gap that produced the error also prevents its recognition.
The propagation problem compounds in classroom settings: a hundred students receiving the wrong answer from a teacher's authority have all updated their understanding in the wrong direction. A student who raised their hand to point out the error would have been doing those students a service, and the institutional culture that discourages students from correcting teachers is directly complicit in the propagation of teacher error.
This mechanism operates at the level of what is taught rather than how student work is evaluated. It takes three forms, all of which share the property that the teacher presents as necessary something that is either unnecessary, redundant, or arbitrary from the student's perspective.
Extra unnecessary steps taught as the procedure. The teacher has learned a method through exposure rather than derivation and has not examined which steps are load-bearing and which are artifacts of the particular presentation they encountered. Students learn the extra steps as part of the method. The schema is encoded with inefficiency built in, and correction later requires identifying and removing steps the student believes are necessary.
Multiple methods taught simultaneously as if each is the simplest approach for a different situation. What are shortcuts for a teacher who understands all of them become competing memory burdens for a student who has mastered none of them. The framework of this pedagogy — documented in the companion papers — is that one derivable method applied to as many situations as possible is superior to multiple memorized methods for specific situations. The teacher who offers multiple methods believes they are being accommodating; the effect on the student is cognitive overload and the false impression that the subject requires remembering many unconnected procedures.
Arbitrary memorization where derivation from fundamentals is available. If a procedure follows necessarily from first principles that have already been established, a teacher who presents it as something to memorize rather than something to derive is encoding the subject as arbitrary rather than logically necessary. This is the most damaging of the three forms because it contradicts the nature of mathematics itself — documented in the companion paper on mathematics as metaphysics (Lacefield, 2026i) — and produces students who experience the subject as a collection of unconnected rules rather than a system of logically necessary relationships that follow from definitions they already hold.
The student has built the concept correctly — the reasoning chain is sound, the approach is appropriate, the understanding is genuine — but a minor computational or notational error has thrown the answer off. The teacher evaluates the wrong answer without examining the reasoning that produced it and marks the work incorrect. The student is told they are wrong when they are, in the most important sense, right.
This is meaningfully different from ordinary error feedback. Ordinary error feedback — telling a student their answer is wrong because their reasoning is wrong — identifies a gap the student can fill. The student can reason more carefully, study the concept, and improve. When a student who reasoned correctly is told their reasoning produced a wrong answer, there is no recoverable path through additional reasoning — because reasoning correctly is already what they were doing. The signal they receive is that their reasoning process is unreliable. That is precisely the cognitive resource that independent mathematical performance requires. Damaging trust in it does not produce improved performance — it produces hesitation, second-guessing, and the learned habit of looking for external permission to trust correct conclusions.
Hattie and Timperley (2007) document that the level at which feedback is delivered determines its effect. Process-level feedback — addressing the reasoning chain — produces substantially different effects from task-level feedback — addressing only the answer. Marking a student wrong at the task level when the error is at the execution level is a mislabeling of the error that sends the wrong signal about where the problem is located.
This is the mirror image of mechanism 3 and operates through the opposite pathway, but the damage mechanism is the same: a mismatch between what the evaluation signals and what is actually occurring in the student's understanding. The student has the correct answer but reached it through wrong reasoning — two added instead of multiplied, arriving at the same result because the numbers happen to make both operations produce the same outcome. The teacher sees the correct answer and gives credit. The student's incorrect reasoning is validated and moves forward as a schema the student now has false confidence in.
This mechanism is underappreciated as a form of incorrect correction because it looks like success. The student got it right. But they got it right for a reason that will fail them the next time the numbers don't cooperate. The schema they carry forward is wrong, and it has been reinforced by an authoritative positive evaluation. This is the accidental-right-answer version of the synthetic model problem documented in the companion paper on the diagnostic — a hybrid of correct output and incorrect structure that feels like understanding and resists correction precisely because it produced correct answers.
The student has the correct answer. They reached it through a method that is logically sound, mathematically valid, and consistent with the reasoning framework that has been established in the course. The method is unfamiliar to the teacher — either because it is genuinely novel or because it was not the method the teacher taught — and the teacher marks it wrong. The student is told they are wrong when they are, in fact, right in the deepest possible sense: they understood the concept well enough to derive an alternative valid approach independently.
This error is categorically different from all others in the taxonomy because it does not merely fail to reward something that should be neutral — it actively penalizes something that should receive the strongest possible positive recognition in mathematics education. The ability to derive a valid method not explicitly taught, using reasoning that was absorbed from prior instruction, is precisely what deep mathematical understanding looks like. It is the goal. Finding it in a student — recognizing it, naming it, celebrating it — is the signal that instruction has succeeded at the level that matters most.
The damage is correspondingly the most severe. A student who derived a valid method, believed they understood, and was told they were completely wrong does not simply lose points. They lose the ability to trust the intuition that they were developing genuine competence. They had said to themselves: I get it, I can do this, I see why it works. The teacher told them that conclusion was wrong. Not that they made an arithmetic error. Not that they needed to show more work. That their mathematical reasoning — the evidence of actual understanding — was wrong. A student who experiences this is being taught that independent mathematical thought is penalized, and that the correct posture toward mathematics is to reproduce what was shown rather than to understand well enough to derive alternatives.
The teacher who makes this error almost always does so not out of malice but out of insufficient subject-matter depth. Ball et al.'s (2008) horizon content knowledge — the awareness of how mathematical topics relate across the curriculum — is specifically what is required to evaluate an unfamiliar method. A teacher without it has no framework for asking whether an approach they haven't seen before might be valid. Their only reference point is whether it matches what they taught. When it doesn't, they mark it wrong.
This mechanism is qualitatively different from the others. It does not operate through feedback evaluation — it operates through deliberate or thoughtless public exposure of a student's gap. A teacher who asks a student a question they do not expect the student to answer correctly, in front of the class, under social pressure, has not made a feedback error. They have used the student's uncertainty as a tool — for classroom management, for emphasis, for their own convenience — without regard for the cost to the student.
The rule is precise: if you are going to call on a student, you should be confident they can answer. If you are not confident, you do not call on them. The purpose of public questioning is to demonstrate understanding, to engage a student who is ready to demonstrate it, or to model reasoning for the class. None of these purposes require putting a student in a public position they cannot handle. A student who is called on and cannot answer has not been taught anything — they have been shown, in front of their peers, that they lack something, in a context that activates fixed-ability attribution rather than productive struggle framing. Dweck's (1988) research is unambiguous on this: public failure experiences in evaluative contexts produce fixed-ability attributions far more reliably than they produce growth motivation.
Board work is a specific and heightened form of this mechanism. When a student is asked to work a problem at the board, the stakes are higher than a verbal question because the work is visible in real time to the entire class. If the student's reasoning goes wrong, their incorrect reasoning is being propagated to every student watching — it becomes the class's experience of how that problem is approached, not only the individual student's. A teacher who allows a student at the board to stray into wrong reasoning without redirecting has allowed one student's confusion to become the class's confusion, which is a distinct and compounding form of damage.
The obligation for board work is therefore higher than for any other public question: only put a student at the board when you are certain they can work through the problem correctly, and be actively ready to redirect — not rescue, not take over, but redirect — the moment their reasoning strays. The redirection should be a question that gets the student back on a valid path, not a correction that signals they were wrong in front of the class.
All six mechanisms operate through the same pathway: they disrupt the student's trust in their own reasoning process — the cognitive resource most necessary for independent mathematical performance, and for academic independence in any subject.
Hattie and Timperley (2007) identify four levels at which feedback can be delivered: task level (the answer), process level (the reasoning), self-regulation level (how the student monitors their own work), and self level (the student's identity and character). Their analysis establishes that feedback delivered at the self level — whether positive ("you're so talented") or negative ("you just don't get this") — produces the smallest learning effects and can actively impede it. The damage mechanisms in this taxonomy all ultimately operate at the self level, regardless of their surface form. The student who is told their correct reasoning produced a wrong answer, or who derived a valid method and was marked wrong, or who was publicly exposed as unable to answer a question they were not ready for, does not receive feedback about a specific task or process. They receive feedback about their competence as a mathematical reasoner. That is self-level feedback, and its effects are correspondingly durable and difficult to repair.
The repair is substantially harder than the damage because it requires rebuilding trust in the reasoning process through accumulated mastery experiences — documented in the companion paper on confidence (Lacefield, 2026b). A single event of incorrect correction can undermine the developing self-efficacy built from multiple prior mastery experiences. Rebuilding it requires multiple subsequent experiences where the reasoning process is trusted, produces correct results, and is recognized as sound — which is a slower accumulation than the single event that disrupted it. The asymmetry between damage rate and repair rate is why these mechanisms are so consequential and why their prevention is categorically more important than their correction after the fact.
"Incorrect correction is often more damaging than the original error. An error can be fixed. A collapse of confidence in one's own reasoning — the resource most necessary for mathematical independence — is substantially harder to repair, because the repair requires trusting the very process that was just penalized."
The practical implication of the full taxonomy reduces to one protocol that prevents most of its mechanisms: ask first, evaluate second.
Before marking any student work as wrong, ask the student to explain their reasoning. This takes approximately thirty seconds and provides the information necessary to distinguish: wrong reasoning (mechanism 3 corrected — address the reasoning), correct reasoning with execution error (mechanism 3 case — affirm the reasoning, correct the execution), correct answer through wrong reasoning (mechanism 4 — address the reasoning), and valid novel method (mechanism 5 — recognize and affirm). Without this step, evaluation at the answer level misidentifies all four cases as equivalent failures and applies the same corrective response — which is wrong in three of the four cases.
The protocol also addresses the board work problem directly. Asking a student to explain their reasoning during board work — before the problem is finished — allows the teacher to detect straying reasoning before it becomes the class's experience. The question "what step are you moving to next and why?" redirects without signaling error, keeps the student on a valid reasoning path, and demonstrates to the class that the reasoning process — not just the final answer — is what the teacher is attending to.
The diagnosis requirement has a time cost. A teacher managing thirty students cannot spend thirty seconds diagnosing every wrong answer in every session. The protocol is most critical for novel concepts, for students who show a pattern of correct reasoning with execution errors, and for any case where a valid novel method might be present. The thirty-second protocol is not a demand that every evaluation be an oral examination — it is a standard that should be applied whenever a student is surprised by a wrong verdict on work they believed was correct.
The subject-matter knowledge argument is strongest in mathematics and formal sciences. In subjects where validity of alternative approaches is more interpretive — some areas of literary analysis, historiography, or philosophical argument — the mechanism 5 damage is real but the protocol for evaluating "valid novel methods" is less precise. The core principle — that a student's reasoning should be examined before their work is evaluated — transfers across subjects, but the specific application requires subject-appropriate judgment.
The self-level feedback claim from Hattie and Timperley (2007) requires precision. Their analysis establishes that feedback delivered at the self level — explicitly addressing the student's identity or capacity — produces poor outcomes. The claim that the mechanisms in this taxonomy function as self-level feedback is an interpretive extension rather than a direct finding of the original paper. The extension is theoretically grounded in Dweck's (1988) work on fixed-ability attribution — public failure experiences in evaluative contexts produce fixed-ability attributions — but should be acknowledged as an inference rather than a direct replication finding.
The six mechanisms documented in this taxonomy share a common property: they are all more common than they are recognized, they occur at every level of education from primary school through doctoral seminars, and they are all substantially preventable through two interventions — sufficient teacher subject-matter depth relative to the level being taught, and the thirty-second diagnostic before evaluation.
The most severe mechanism — penalizing a valid novel method — deserves explicit reiteration as the paper's strongest claim. It is not merely an error. It is a systematic inversion of the reward structure that mathematics education is supposed to produce. A student who derives a valid method independently has demonstrated exactly what deep mathematical understanding looks like. That this achievement is regularly penalized — because the teacher lacks the subject-matter depth to recognize validity in an unfamiliar method — is not a minor flaw in instructional practice. It is a failure at the level of the educational enterprise's stated purpose.
Mathematics is used throughout this paper as the primary worked example because its precision makes the mechanisms most demonstrable. The framework applies across all subjects in which reasoning can be examined, in which teacher subject-matter knowledge has a genuine depth requirement, and in which the difference between right-answer-wrong-reasoning and wrong-answer-right-reasoning is educationally consequential. That description covers most of what is taught in most educational settings.