## A sentence can arrive before the first question
Some students sit down in mathematics already carrying an answer. Not an answer to the problem on the board. An answer to a deeper question: *What kind of person am I when mathematics appears?*
“I’m not a math person.”
The sentence can come from a learner who is genuinely struggling. It can also come from a learner whose results are respectable, whose errors are repairable and whose reasoning is stronger than they realise. Sometimes the student says it after a difficult test. Sometimes they say it before trying. Sometimes it has been repeated so often that it has become an identity statement rather than a description of current performance.
That distinction matters.
A student can have a weak topic and still possess a healthy mathematical identity: “I do not understand simultaneous equations yet, but I know what to do when I am stuck.” Another student can score well while carrying a fragile identity: “I am only good at mathematics when the questions look familiar. If I cannot see the method immediately, it proves I was never really good.”
The first learner has a problem to solve. The second learner has a self-story that turns ordinary difficulty into evidence about who they are.
Current education publishing is paying renewed attention to this issue. Edutopia published a 2026 piece specifically on building a healthy math identity and, in September 2026, another on secondary students who repeatedly say they are “not a math person.” The theme is not cosmetic. Identity changes how learners interpret error, speed, comparison, help, challenge and success. It shapes whether a hard question feels like a solvable problem or a verdict.
The purpose of this article is not to suggest that confidence alone produces mathematical skill. It does not. Mathematics still requires knowledge, representations, procedures, concepts, practice and reasoning. The more useful claim is narrower: **students learn mathematics through a loop in which performance affects identity, and identity affects the next performance.**
If we want to improve the mathematics, we should understand the loop.
## 1. Math identity is the story a learner extracts from repeated experiences
Identity sounds abstract until we define it operationally. In this context, mathematical identity is a learner’s running answer to questions such as:
– Do people like me belong in mathematics?
– What counts as being “good” at it?
– When I struggle, what does that struggle mean?
– Is speed a sign of intelligence?
– Are mistakes evidence of low ability or information about the next move?
– Do my explanations matter, or only the final answer?
– Am I someone who can recover when the first method fails?
Students rarely sit down and answer these questions formally. They infer the answers from classroom life.
A pupil who is praised only for being fast may conclude that mathematical ability is speed. A pupil whose correct but unusual method is ignored may learn that mathematics means reproducing the teacher’s route. A learner who gets help only after public failure may associate asking questions with exposure. A student who repeatedly receives easier work may interpret support as evidence that adults do not expect much from them. A high-performing student praised as “naturally gifted” may become terrified of the first topic that does not come easily.
Identity is therefore not a motivational poster inside the child. It is partly a record of experience.
This is good news because experience can change.
## 2. “Not a math person” can mean several different things
Teachers should resist answering the sentence too quickly. “Of course you are!” is kind, but it may miss the diagnosis.
When a learner says “I’m not a math person”, they may mean:
**I have missing prerequisite knowledge.** Fractions, algebraic manipulation or number sense may be unstable, so each new topic feels harder than it should.
**I am slow and I have learned that slow means bad.** The learner may reason well but compare their speed with classmates who answer first.
**I can follow examples but cannot start unfamiliar problems.** The issue is method selection, not basic understanding.
**I panic when I am evaluated.** This is closer to test anxiety, where threat consumes attention and working memory.
**I have had years of public correction.** Mathematics has become socially risky.
**I succeed only when I can imitate.** The student senses that their performance is brittle and interprets that brittleness as lack of talent.
**I do not see myself represented in who is treated as a mathematician.** Belonging and stereotype signals can shape participation.
**I dislike the subject.** Preference and capability are not the same, but students often merge them.
Each cause suggests a different response. Identity work without diagnosis can become empty encouragement. Diagnosis without identity work can repair the skill while leaving the learner afraid to use it.
The right question is: **What evidence taught this student to say this sentence?**
## 3. Speed is one of the most powerful accidental identity signals
Mathematics classrooms make speed visible. A hand rises. A page fills. A student announces an answer. Someone finishes first. Timed drills, mental calculations and rapid recall all have legitimate uses, but they can also create an unintended hierarchy: fast equals smart.
That hierarchy is misleading.
Some mathematical work should become fluent. If every basic fact requires a long reconstruction, working memory becomes overloaded during multi-step problems. But fluency is not identical to mathematical intelligence. Complex mathematics often rewards representation, persistence, checking, structural noticing, proof, approximation and the willingness to abandon an unproductive route.
A student who takes two minutes to notice that a problem has hidden symmetry may be doing more valuable mathematical thinking than a student who launches into thirty seconds of unproductive algebra.
The identity problem appears when speed is treated as a global trait. “She is fast” becomes “she is good at maths.” “He needs time” becomes “he is weak.” Students absorb the classification.
A better classroom makes different forms of competence visible. Sometimes ask: Who found a second method? Who spotted an assumption? Who corrected a sign error before it spread? Who can explain why the answer is reasonable? Who can show where another method would fail? Who asked the question that unlocked the problem?
The message is not “speed never matters.” The message is “speed is one mathematical property among many.”
## 4. High-achieving students can have especially fragile identities
We often assume math identity is mainly a problem for students with low marks. High achievers can be just as vulnerable, in a different way.
Suppose a student has been first in class for years. Adults repeatedly say, “You are brilliant at maths.” The praise feels positive, and for a while it may be. But the student may quietly build a conditional identity:
**I am a math person because I get things quickly.**
Then Additional Mathematics, olympiad-style reasoning, a new school, a stronger peer group or one difficult examination arrives. For the first time, the student does not understand immediately. If difficulty has never been part of their identity model, the experience is not interpreted as “this is a harder problem.” It becomes “maybe I was never actually good.”
This is why a healthy identity cannot depend only on results. It needs process evidence.
A stronger self-description is: “I know how to enter a hard problem. I can represent it, test a simpler case, identify what is known, retrieve related methods, check my working, ask a precise question and learn from feedback.”
That identity survives a bad mark better because it is anchored in controllable mathematical actions.
## 5. Anxiety and identity overlap, but they are not the same thing
Math anxiety is a real and important phenomenon, but it should not be used as a catch-all explanation.
An anxious learner may experience physiological arousal, intrusive worry or working-memory interference when mathematics is performed under pressure. Identity asks a different question: what does the learner believe the difficulty says about them?
They interact. A student who believes “people like me are bad at mathematics” may interpret a racing heart as confirmation. A student who experiences repeated anxiety may conclude that mathematics is fundamentally incompatible with them.
But the distinction matters for intervention. Breathing strategies, test familiarisation and pressure management may help anxiety. They will not repair a missing concept. More practice may repair fluency but not necessarily change a student’s belief that only fast students belong. Encouragement may improve mood but not fix an unstable fraction model.
A useful diagnostic sequence is:
**Skill:** What can the learner actually do?
**State:** What happens under pressure?
**Story:** What meaning does the learner attach to success and failure?
**System:** What classroom experiences keep reinforcing that story?
Treating these as separate layers makes support more precise.
## 6. Mathematical agency is a better target than generic confidence
Confidence is attractive because it sounds positive. But confidence can be inaccurate. A student can feel confident and be wrong; another can feel uncertain and reason well.
A stronger target is **mathematical agency**: the learner’s capacity to act intelligently when the route is not obvious.
Agency sounds like:
– I can start somewhere sensible.
– I can choose a representation.
– I can test whether my answer is plausible.
– I can notice when my method is failing.
– I can ask for a hint without surrendering the whole problem.
– I can compare two methods.
– I can recover from an error.
– I can explain what I know and what I do not know yet.
Agency is evidence-based. It grows from repeated experiences of making a move and seeing that move improve the situation.
This is especially important for students who have been told to “believe in themselves”. Belief unsupported by capability is fragile. Capability that the learner can recognise becomes durable confidence.
The sequence should often be **action → evidence → belief**, not belief first and evidence later.
## 7. Begin identity repair with tasks that reveal competence honestly
If a student has years of negative mathematics experiences, giving them an extremely easy worksheet can backfire. They may read the simplification as proof that the teacher agrees they are incapable.
The challenge is to create **accessible success without fake success**.
Low-floor, high-ceiling tasks are useful here. Everyone can enter, but the mathematics can deepen. A number pattern, visual proof, estimation problem, multiple-solution puzzle or “which one does not belong?” prompt can let a learner contribute before formal technique becomes the only currency.
The teacher should then name the competence accurately.
Instead of “See, you are smart!”, say: “You noticed the pattern before we had an algebraic rule.” Or: “You checked the answer against the diagram and caught a contradiction.” Or: “You changed representation when the first one stopped helping.”
Specific feedback gives the student a new data point about themselves.
Identity changes slowly because one positive lesson must compete with years of evidence. That is normal. The objective is not to force a student to say “I love maths.” It is to accumulate credible experiences that make the old story less accurate.
## 8. Let students see mathematics as decisions, not answer production
Students often see expert work only after it has been cleaned up. The teacher presents the efficient solution; the textbook prints the polished method; the mark scheme rewards the final structure. The discarded attempts disappear.
This can make mathematics look like a performance where good students instantly know what to do.
One of the most powerful identity interventions is to make expert decision-making visible.
A teacher can model:
“I could expand this expression now, but that may create more terms. I’m going to pause and look for structure.”
“I expected this answer to be positive. It is negative, so I need to inspect the sign rather than continue.”
“I tried substitution and it became messy. That is information. I am switching representation.”
“I do not know the exact route yet, but I know what I can establish first.”
This does not lower mathematical rigour. It exposes the rigour that polished solutions hide.
When students see competent adults pause, check and revise, difficulty stops being incompatible with expertise.
## 9. Error analysis should protect the learner while exposing the mathematics
“Celebrate mistakes” is a popular phrase, but students know when adults are performing optimism. Some mistakes feel embarrassing. Some cost marks. Some reveal genuine gaps. We do not need to pretend errors are pleasant.
We need to make them useful.
A productive error routine separates **person**, **error** and **repair**.
Person: the learner is not the error.
Error: what exactly went wrong?
Repair: what rule, representation, check or habit would prevent recurrence?
Suppose a student repeatedly writes \((a+b)^2=a^2+b^2\). The identity-damaging response is global: “You are careless with algebra.” The mathematically useful response is structural: expand the brackets, compare with an area model, identify the missing cross-term, then create a self-check.
The learner should leave with a repair tool, not merely a red mark.
For older students, build an error log that records categories rather than shame: misread condition, wrong base, sign propagation, unjustified cancellation, method-selection failure, calculator entry, interpretation error. Patterns turn failure into diagnostic information.
Agency grows when a learner can say, “I know the kind of mistake I make, and I have a way to catch it.”
## 10. Representation can change identity because it changes entry points
A student can feel weak in symbolic mathematics while thinking strongly through diagrams, tables or concrete quantities. If the classroom presents only one representation, the learner may confuse difficulty with that representation for inability in the underlying idea.
Good teaching moves between forms:
– concrete quantities;
– diagrams and models;
– tables;
– graphs;
– symbols;
– equations;
– verbal explanations;
– numerical examples.
This is not about assigning learners fixed “styles”. It is about matching representation to the mathematical structure and using translation between representations as a learning tool.
For example, simultaneous equations can be symbols, intersecting lines or two conditions that must be satisfied at once. Fractions can be division, ratio, measure, operator or part-whole relation. A function can be a rule, mapping, graph, table or relationship between quantities.
When a learner finds one doorway into the idea, they can build outward. The message becomes: “There is a route into this problem,” not “I either see it instantly or I do not belong.”
## 11. Student-generated problems turn learners from recipients into authors
A strong mathematical identity includes ownership. One way to build it is to let students create, vary or repair problems.
After solving a linear equation, ask students to design another equation with the same solution but a different structure. After a percentage problem, ask them to create a scenario where choosing the wrong base would produce a tempting error. After a geometry lesson, ask them to draw a figure that satisfies three constraints and challenge a partner to find an unknown angle.
Problem creation changes the role of the student. They are no longer only being judged by mathematics; they are using mathematics to design something that can judge someone else’s reasoning.
It also reveals understanding. A student who can generate a valid counterexample, construct a solvable problem or vary one condition while preserving another is showing structural knowledge.
Edutopia highlighted student-created word problems in 2026 as a way to build ownership, collaboration and deeper learning. The broader principle is valuable: **authorship changes the learner’s relationship with the subject.**
## 12. Language matters: stop turning temporary states into permanent identities
Adults often use trait language casually.
“She’s a maths kid.”
“He’s just not mathematical.”
“She’s careless.”
“He’s naturally gifted.”
These phrases compress complex performance into identity labels. Even positive labels can create pressure. If a child is “the clever one”, a hard task threatens more than a grade.
Better language describes observable states and strategies:
“You currently solve routine equations accurately, but unfamiliar setup is still difficult.”
“You are quick at calculation; now we are working on explanation and proof.”
“You understand the concept when it is drawn. Let’s connect that to the symbols.”
“You lost marks from sign propagation, not because you did not understand the method.”
This language keeps the system repairable. It tells the learner what changed, what has not changed and what the next job is.
## 13. Comparisons should be engineered carefully
Students compare themselves even when teachers do not ask them to. Who finished first? Who got the highest score? Who was asked the hard extension? Who needs extra help?
Comparison is not always harmful. It can help a learner calibrate performance and see models of strong work. The problem is when comparison becomes identity ranking.
Classrooms can redirect attention toward **within-person progress** and **strategy comparison**.
Instead of only displaying the top score, compare a student’s first and second attempts. Instead of asking “Who got it right?”, ask “What different methods appeared?” Instead of ranking speed, analyse why one method is efficient under certain conditions.
For assessments, show improvement by error category. A student may still have an average overall score while having eliminated one recurring misconception. That is important evidence of capability growth.
Identity becomes healthier when students can see themselves changing.
## 14. Help-seeking should be treated as a mathematical skill
Some students think asking for help proves they are weak. Others ask for help so early that they never experience productive struggle. Both patterns reduce agency.
Teach **graduated help-seeking**.
Before asking, the learner should identify:
1. What is the question asking?
2. What information do I have?
3. What have I tried?
4. Where exactly did the route stop?
5. What is the smallest hint that would let me continue?
This changes “I don’t get it” into a precise mathematical request.
A tutor or teacher can respond with the smallest useful intervention: point to a representation, ask a question, remind the learner of a related case, reveal one step or model a decision. The goal is to return control quickly.
The learner should then finish enough of the route to own it.
This matters for identity because students begin to experience help as a tool they can direct, not rescue that confirms dependence.
## 15. Families can either freeze or loosen mathematical identity
Parents often try to comfort children with statements such as, “Don’t worry, I was never good at maths either.” The intention is empathy. The unintended message can be inheritance: perhaps mathematics ability is a family trait you either receive or do not.
A better response keeps difficulty specific.
“What part is difficult?”
“Show me what you know before the point where it breaks.”
“What did the teacher’s example do differently?”
“How could you check that answer?”
“What did you learn from the last attempt?”
Parents also need to separate marks from identity. A 55 is information about one assessment under specific conditions. It may reveal missing knowledge, weak timing, careless execution, anxiety, unfamiliarity or several of these together. It does not reveal the child’s fixed mathematical ceiling.
At the same time, avoid false reassurance. “You’re amazing at maths” is less useful than “You used to avoid ratio problems; now you can set up the base correctly and explain why.” Evidence is more convincing than slogans.
## 16. A five-part mathematics identity audit
Teachers and tutors can use a simple audit when a student’s relationship with mathematics seems to be limiting performance.
### A. Capability
What knowledge and procedures are actually secure? Which prerequisites are missing? Can the student perform with and without examples?
### B. Entry
Can the student begin an unfamiliar problem? Do they know how to represent it, simplify it or identify constraints?
### C. Recovery
What happens after an error? Does the learner inspect, switch, check, ask or shut down?
### D. Interpretation
What meaning does the student assign to speed, mistakes, help and comparison?
### E. Evidence
What recent experiences could provide credible counter-evidence to an unhelpful identity story?
This audit prevents a common mistake: treating a learner’s statement as either purely emotional or purely academic. Usually it is connected to both.
## 17. What teachers should not do
Several well-meant approaches can make identity work weaker.
**Do not replace mathematics with motivation.** Students eventually know whether they can solve the problem. Skill must improve.
**Do not remove all difficulty.** A learner cannot build an identity around recovery if they are never allowed to struggle safely.
**Do not praise fixed traits.** “You are a genius” can make future difficulty threatening.
**Do not publicly rank students by speed.** Fluency has a place; identity hierarchies do not need to be built around it.
**Do not confuse a calm student with a confident one.** Some students conceal anxiety well.
**Do not assume high marks mean healthy identity.** Perfectionism and fear of failure can coexist with excellent performance.
**Do not assume low marks mean low potential.** Diagnose the mechanism before narrating the learner.
## 18. What progress actually looks like
A healthier mathematical identity does not necessarily sound like “I love maths now.” It often appears in quieter behaviours.
The student starts before asking for help.
They annotate what is known.
They try a second representation.
They can name the exact point of confusion.
They stop treating every error as evidence of inability.
They check an answer without being told.
They can watch a faster classmate without deciding the comparison settles their own potential.
They are willing to show incomplete working.
They recover after a difficult paper.
They choose a challenging question occasionally because challenge has stopped being a threat to identity.
These are signs of agency. Marks may improve later, sometimes gradually. The behavioural shift is important because it changes the amount and quality of mathematical practice the learner is willing to do.
## 19. Small-group mathematics can change the evidence a learner receives about themselves
Mathematical identity is partly social. In a large room, a learner can disappear behind the fastest voices, the most confident volunteers or the quiet safety of copying. A well-run small group changes the information available to both teacher and student. Thinking becomes more visible. The teacher can see whether a student is unable to begin, chooses a weak representation, understands the idea but loses notation, or knows the method and simply needs more processing time.
That visibility matters because identity is often damaged by inaccurate conclusions. A student who needs ten extra seconds may have been labelled weak when the real issue is processing pace. A learner who produces the right answer may have been labelled secure even though they cannot explain why it works. A quiet student may possess a strong method that never enters the public conversation.
Small groups are not automatically better. They can also intensify comparison if one learner dominates and the others become spectators. The teacher has to design participation. One student might explain the representation, another identify a constraint, another check the result, and another propose an alternative route. Roles should rotate so that identity does not harden into “the fast one”, “the checker”, “the weak one” and “the quiet one”.
The most useful small-group feedback is granular. “You are getting better at algebra” is vague. “Last month you often expanded before checking structure; today you factored first and avoided six unnecessary lines” gives the learner something real to own. “You asked for a hint only after identifying the exact step that broke” gives evidence of improved help-seeking. “You caught your own domain error before I said anything” gives evidence of self-monitoring.
Over time, these observations accumulate into a more accurate self-model. The learner begins to see that mathematical competence is not a single score. It is a system of habits, representations, knowledge and decisions that can be improved one component at a time.
## 20. Assessment can either freeze identity or turn it into a learning signal
Tests are powerful identity events because they compress weeks of learning into a number. The number is useful, but students often overinterpret it. A 62 can become “I am a 62-percent maths student.” A distinction can become “I must always be the distinction student.” Both interpretations are too large for the evidence.
A better assessment conversation separates at least four things: coverage, correctness, robustness and transfer.
**Coverage** asks which content appeared and which did not. A test never samples the entire subject.
**Correctness** asks what the learner could execute accurately under those conditions.
**Robustness** asks whether the method survives pressure, time and multi-step dependency.
**Transfer** asks whether the learner can use the idea when the surface form changes.
Two students with the same mark may have very different profiles. One may know nearly everything but lose marks to time and checking. Another may complete familiar procedures cleanly but fail novel applications. A third may have one large prerequisite gap that contaminates several topics. If all three receive the same message—“work harder”—assessment has failed as diagnosis.
Post-test review should therefore produce a small action map. Which errors were knowledge gaps? Which were representation choices? Which were execution slips? Which were reading errors? Which were time-allocation failures? Which answers were lucky because the working could not justify them? Which wrong answers contained a sound approach worth preserving?
This approach also protects high achievers. A student who falls from 90 to 75 does not need a dramatic identity crisis if the analysis shows that one new topic and a time-control problem explain most of the loss. Conversely, a student who rises from 55 to 65 can see meaningful progress if three old error categories have disappeared even though the overall grade is not yet where they want it.
Assessment should narrow the next job. When a test instead becomes a global judgement of the learner, it creates exactly the kind of fixed story that makes future mathematics harder.
## 21. Three learner profiles show why identity repair must be precise
Consider three students who all say, “I’m not a math person.” The sentence is identical; the mechanisms are not.
**Learner A: accurate but slow.** Aisha usually reaches correct answers, but she takes longer than classmates. She watches hands rise while she is still organising the problem and concludes that real mathematicians “just see it”. Giving her more easy work will not help. She needs a classroom where processing time is protected, where good checking and multiple methods are visible, and where fluency is developed without treating speed as a global intelligence score. A useful goal might be faster retrieval of a few foundational facts while preserving the careful reasoning that is already a strength.
**Learner B: fast but brittle.** Ben scores well on familiar exercises and has long been praised for natural talent. When a problem changes form, he freezes. Because his identity depends on effortless success, he avoids challenge and interprets asking for help as exposure. His repair is almost the opposite of Aisha’s. He needs unfamiliar problems where initial uncertainty is normal, explicit modelling of method selection, and feedback that values recovery rather than first-attempt speed. The teacher may deliberately show expert false starts so difficulty stops meaning “not gifted”.
**Learner C: persistent but structurally blocked.** Chen works hard yet repeatedly fails algebra because fraction operations are unstable. Adults have encouraged him for months, but each new algebra lesson produces the same breakdown. Identity work without skill repair would be unfair. He needs a bounded prerequisite intervention that rebuilds fraction structure, then immediate opportunities to use the repaired knowledge inside algebra. His identity changes when the world gives him new evidence: problems that used to collapse now remain solvable.
These profiles illustrate a core rule: never let a motivational label replace a mathematical diagnosis. “Low confidence” is not specific enough. Neither is “weak at maths”. Ask where the chain breaks and what story the learner has built around that break.
The teacher can then choose a paired intervention: one move for capability and one for interpretation. For Aisha, perhaps fluency practice plus protected wait time. For Ben, unfamiliar problem solving plus praise for revision and persistence. For Chen, prerequisite repair plus visible tracking of the errors that disappear.
Identity becomes healthier when the student receives accurate evidence that the system is changing. That is much more durable than asking every learner to repeat the same positive sentence about mathematics.
## 22. Homework can reveal whether identity is becoming more independent
Homework is often discussed only as practice volume, but it also shows how a learner behaves when the immediate teacher is absent. A student with developing agency begins to leave useful traces: a diagram before an equation, a note beside a failed method, a question mark at the exact point of uncertainty, a check beside a verified answer. Those traces are evidence that the learner is managing the mathematics rather than merely waiting for rescue.
This changes how homework should be reviewed. Instead of marking only completed answers, teachers can occasionally reward the quality of the attempt. Which question did the student return to? Where did they identify a contradiction? Did they copy a worked example mechanically, or adapt its structure? Did they stop after the first wrong answer, or test another route?
For families, this also gives a boundary. Helping does not have to mean solving. A parent can ask, “What have you tried?” and “Where exactly did it stop making sense?” If the child can name the blockage, they have already done part of the mathematical work.
Over time, independent homework becomes one more source of identity evidence. The learner sees that mathematical progress is not something that happens only when a strong adult is sitting beside them. They can create the next move themselves.
## Conclusion: the goal is not to manufacture confidence; it is to build a learner who can act
“I’m not a math person” sounds like a statement about ability. Often it is a compressed history of classroom experiences, comparisons, errors, praise, anxiety, missing knowledge and beliefs about what mathematics is supposed to feel like.
The response should not be denial. It should be diagnosis and reconstruction.
Repair prerequisite knowledge. Make more forms of competence visible. Separate speed from intelligence. Model expert uncertainty. Turn errors into specific repair jobs. Give students real authorship. Teach precise help-seeking. Let them see evidence of their own change.
Most importantly, build mathematical agency.
A learner does not need to feel certain before every problem. They need to know how to move when certainty is absent.
That is a stronger identity than “I am naturally good at maths.” It can survive a difficult chapter, a disappointing test and a room full of faster hands.
The aim is a student who can say something more useful:
**I may not know the route yet, but I know how mathematicians begin.**
—
## Related eduKateSG reading
– **How Student Confidence Works | Evidence Before Belief**
How Student Confidence Works | Evidence Before Belief
– **How Test Anxiety Works | When Threat Competes With Performance**
How Test Anxiety Works | When Threat Competes With Performance
– **How Test Anxiety Affects Performance | When the Alarm System Enters the Examination**
How Test Anxiety Affects Performance | When the Alarm System Enters the Examination
– **How Attributional Retraining Works | Change the Explanation of Failure So the Next Strategy Can Change**
How Attributional Retraining Works | Change the Explanation of Failure So the Next Strategy Can Change
## Sources and further reading
– Edutopia (22 January 2026), *How to Build a Healthy Math Identity*.
https://www.edutopia.org/video/how-to-build-a-healthy-math-identity/
– Edutopia (10 September 2026), *Building Students’ Confidence and Understanding in Math*.
https://www.edutopia.org/article/supporting-secondary-students-building-confidence-math
– Edutopia (3 August 2026), *Setting Up Math Activities That Are Connected to Students’ Lives*.
https://www.edutopia.org/article/setting-up-math-activities-that-are-connected-to-students-lives
– Edutopia (30 January 2026), *Student-Created Math Word Problems Help Motivate Deep Learning*.
https://www.edutopia.org/article/student-created-math-word-problems/
