Homeschool practice

What to do when your homeschooler stalls on math drills

Written 10 September 2026 by Daniel, who built this

A homeschooler who has stopped doing math drills is rarely telling you the curriculum is wrong. It’s usually one of three things: the work sits above where the kid actually is, the sessions run too long, or the kid has decided they’re bad at math and worked out that refusing is cheaper than being proved right. All three are fixable this week, and none of them requires buying anything.

The curriculum switch is the expensive answer

When a kid digs in, the materials are the easiest thing to blame, partly because they’re the one part of this you could replace by Friday. I have felt that pull. But a new program teaches the same facts in a different order with better packaging, and the stall walks straight across into it. Work through the three below first. If you have honestly done all three and nothing has moved, switching becomes a decision rather than a guess.

A stall is usually one of three things

The work sits above where the kid actually is

This is the common one, and hard to see from across the table, because the level looks correct on paper. A ten year old is supposed to be on the times tables, so the times tables are on the sheet. But if the twos, fives and tens aren’t sitting in memory yet, every row is a small piece of arithmetic built from scratch, and forty rows of that’s exhausting in a way that has nothing to do with attitude.

The test I use is the pause. Sit next to them and watch the gap between reading the question and the pencil moving. A fact that’s known produces no gap. A fact being rebuilt produces a visible one, usually with fingers or lips involved. If more than a handful of questions on a page produce that pause, the page isn’t practice. It’s a test, and they’re failing it daily at the kitchen table.

The fix is uncomfortable, which is why it gets skipped. Drop the level until the work is genuinely, boringly easy. Easy enough that they finish and look at you as though you’re wasting their time. Stay there for a week or two. Fluency gets built on top of facts that are already automatic, so this isn’t going backwards. It’s laying the part that was missing.

The session runs too long

Ask a kid for fifty problems and watch where the accuracy goes. The last third is where the wheels come off, and the last third is also the part that gets remembered: the eraser marks, the sighing, the argument about whether they can stop. A session that ends badly teaches a kid that math sessions end badly, whatever the first thirty problems looked like.

Short and daily beats long and twice a week, and it’s not close. Five minutes that get finished are worth more than twenty that get abandoned, because what you’re training at this stage is the sitting down, not the arithmetic. Ten questions for a younger child and twenty for an older one is the range I settled on.

The kid has decided they’re bad at math

This one doesn’t respond to changing the worksheet. A kid who has concluded they’re bad at math is protecting themselves, and refusing to start is an effective way to do it. Starting and getting things wrong produces evidence, and they believe they know what it will say.

One piece of research made me take this more seriously. Ramirez, Gunderson, Levine and Beilock tested 154 first and second graders and found the expected negative relationship between math anxiety and math achievement, but only among the children with higher working memory. Their argument is that anxiety occupies the same mental workspace the better strategies run in, so the kids with the most to lose lose the most: Math Anxiety, Working Memory, and Math Achievement in Early Elementary School.

I want to be careful with it. That’s a correlational study of young children in classrooms, not of homeschoolers and not of timed drills, so I’m carrying a finding across rather than citing a study of this exact thing. What it changed for me is the assumption underneath the problem. A kid stalling on math facts isn’t evidence of a kid who is bad at math. It can be evidence of a kid who is worried, and worry is loud enough to look like inability.

The practical change is what you say out loud. Stop letting the drill be evidence in either direction: praising a good round confirms the same frame a bad round does, which is that the round is a verdict. Comment on the fact that it happened instead. That’s five days this week. The score can stay between the kid and the page.

Keep the drill and the teaching apart

The change that made the most difference here was refusing to teach anything new during drill time. A fact drill should hold no surprises. When a new idea turns up inside one, a borrow they have not seen or a remainder or a fraction, the drill has become unpredictable, and unpredictable is the thing a stalling kid is avoiding.

So teaching gets its own block, with a whiteboard, no clock, and however long it takes. Drill is separate, short, fast, and old ground only. Only one of them is allowed to be hard, which also gives you somewhere to put a bad day: a concept that’s not landing is a teaching problem, not a reason to cancel drill.

Where paying for the round helps, and where it doesn’t

I built PiggyBankMath for my two kids, so read this as interested rather than neutral.

It’s useful for exactly one of the three problems above: the habit. A kid gets paid for finishing a round, never for the score, and finishing is something they can decide to do on a bad morning. A round is ten questions for a seven year old, fifteen through age ten and twenty for older kids, and you can override that if it’s still too much. You set what a round is worth and a daily maximum, which starts at a dollar. When cash changes hands you press Mark as Paid. The app processes no payments and holds no money; it holds the record.

Two details matter more than the money. When the day's maximum is reached the money stops and the practice doesn’t, so a kid who keeps going is never told the work stopped counting. And the optional five minute assessment places them at their real level, which is the first problem on this page.

What it doesn’t do is worth stating plainly. It can’t teach a concept. If a kid doesn’t understand what division is, a quarter won’t explain it, and a round of division facts is only a faster route to feeling bad. It doesn’t solve a level problem by itself either: the assessment places them, but you’re still the one who decides the work is too hard. And for a kid anxious specifically about being timed, a timed round may be the wrong tool this month. Each kid has a minimum accuracy setting, and a round below it earns nothing and says so on screen. I would set that low here, or leave the whole thing alone until the fear comes down.

The part I’m not sure about

Three reasons is a tidy number and real kids aren’t tidy. A stall can be about a sibling who finishes faster, or a bad couple of weeks that have nothing to do with math, or the fact that drill sits right after lunch. I have two kids and a lot of reading, not a study. If you work through the level, the length and the framing and nothing moves, ask the kid directly, at a time that’s not math time. That conversation has told me more than any change I ever made to a worksheet.