Math homework with ADHD: five traps and four moves.

Before the playbook.

Two things up front. First: a child with ADHD is not lazy, not unmotivated, and not "choosing" not to focus. The executive-function machinery that lets a kid ignore an itch and finish a worksheet runs on a different rhythm in an ADHD brain — slower to start, faster to disengage, more sensitive to the gap between "this is interesting" and "this is hard." That's not a willpower problem. It's an architecture problem. The moves in this post are about working with that architecture, not pushing against it.

Second: we're not the parent in the room. We're the team building a Pre-K-8, multi-subject tutor for kids — launching with math for grades 2 to 5 — and we've spent the last year reading a lot, talking to families, and watching kids work. What follows is what we keep seeing — both the patterns that make the homework battle worse and the small moves that make it less of a battle. None of this is a substitute for an evaluation, an IEP, or a real plan with your child's teacher. Treat it as field notes from people who care about getting this right.

Five traps that make it worse.

1. "Just try harder."

ADHD doesn't respond to effort the way the homework instructions assume. If your child could "just focus," they would have done that already. Saying it doesn't add focus; it adds shame, and shame is the single most reliable way to make the next 10 minutes worse. The replacement isn't a different motivational line; it's a different question: "What part is hard right now?" The answer is often very specific (the hand hurts, the page is too crowded, the third problem looks like a wall) and points at a fix that "try harder" doesn't.

2. The 30-page packet.

A standard homework packet for a 4th-grader with no other obligations is a lot. The same packet for an ADHD 4th-grader, who's already spent 6 hours managing attention at school, is a 90-minute commitment of the kind of focus they have least of by 4 p.m. The trap: making the packet visible all at once. The kid sees 30 problems, can't model finishing 30 problems, and shuts down before the first. The fix is small — fold the packet in half, or cover the bottom with a piece of paper, or print four problems at a time. The math doesn't change; the visible commitment does.

3. Reading the answer out loud.

When your kid is stuck on 7 × 8 and you say "fifty-six," you've ended the problem. It's tempting because it's fast and the kitchen is full of dinner. But the long-run cost is that your child learns to wait for the answer instead of working through it, and the next problem produces the same stuck moment. What works better is offering one rung of a ladder: "You know seven sevens — what's one more seven?" That's still you helping. It just doesn't end the problem.

4. Tablets that go ding.

Most ed-tech apps are built to capture attention. They make a sound when your kid gets a problem right; they reveal a new costume; they show streaks and badges and a friendly cartoon character clapping. For a kid without ADHD, this is a small motivator. For a kid with ADHD, it's a slot machine. The dopamine spike from the costume drowns out the satisfaction of having understood something, and over weeks the satisfaction of understanding stops being the thing the kid is chasing. We've watched this pattern in our own families and in the literature on variable reward schedules; it's one of the reasons Koda doesn't ship with streaks on by default.

5. "Don't break your streak."

Streaks deserve their own line. They are an excellent retention metric for an app and an excellent way to convert a child's relationship with math from "I am learning" to "I am protecting a number." The first time the streak breaks — flu, vacation, a bad day — the loss feels real, and the kid often quits. ADHD kids are especially susceptible because the ratio of quit moments to recommit moments is higher when starting is hard. If your child's math app has a streak, see if you can turn it off. If you can't, see if you can find a different app.

Four moves that make it better.

Move 1 — Pick a focus chunk that fits this kid, today.

For most ADHD 4th- and 5th-graders, 25 minutes is too long; 5 minutes is too short to settle into. The sweet spot is somewhere between 10 and 18 minutes. This kid, today is the operative phrase — a child who could do 18 yesterday might only have 12 today, and pretending otherwise is the start of a fight. Set a visible timer. When the timer goes off, the chunk ends, even if there's one problem left. The kid gets a real break (water, walk, fidget, view out the window). Then back, on the kid's terms.

Move 2 — Concrete first, always.

The standard sequence for any new math idea is: concrete, then pictorial, then symbolic. ADHD kids particularly benefit from staying in the concrete phase longer — counting bears, fraction circles, base-ten blocks, a number line drawn with a pencil — before the algorithm. The mistake to avoid is jumping to the algorithm because it's faster on paper. It's not faster if your child loses the thread. It's just faster on the day they already understood it.

Move 3 — Tag the effort, not the answer.

"You showed every step" is more useful than "you got it right." For a kid whose self-talk has slid into "I'm bad at math," the second one is fragile (next time they're wrong, they're right back to bad-at-math) and the first one is durable (showing the steps is a behavior, not a verdict). This isn't praise theater — it's just naming the part you actually want more of. Most ADHD kids are working harder than the right answer suggests, because every right answer cost them five times as much regulation as their classmates'. Naming the effort lets the kid see the effort.

Move 4 — Warn the transitions.

"Two more problems and then a break" is more useful than the timer alone, because the timer is invisible until it goes off and the transition is the part ADHD kids find hardest. A 2-minute warning before any shift — break starting, break ending, dinner, bedtime, the next subject — costs almost nothing and replaces a fight with a heads-up. Pair it with a visible cue (a card, a finger on the watch, a note on the page) and your child can see the transition coming, which is most of what makes it survivable.

One thing that helps but doesn't have to.

Body co-regulation. Sit at the table with your child while they work, even if you're not doing anything related to the homework. Read a book. Pay your bills. Drink your coffee. The presence of a calm adult in the same room — not hovering, not asking questions, just present — does something for an ADHD kid's regulation that being alone in their bedroom doesn't. It isn't a guarantee, and it isn't always possible (every parent has work, dinner, other kids), but it costs nothing on the days you can manage it and it tends to make the chunk land better.

What we built.

We're writing this in part because it's the design brief for Koda. The four moves above aren't add-ons in the product; they're the defaults.

• Focus chunks are configurable from 5 to 25 minutes per child, and the timer is visible the whole time. When it ends, Koda suggests a real break and waits.
• Most concepts ship with a visual or hands-on explainer (base-ten blocks, fraction circles, area grids, number lines) alongside the standard algorithm. Your child picks whichever angle has rendered for that topic. (188 short videos in the catalog, drawing on five teaching angles — concrete, number line, area, standard algorithm, word problem; software updates add more per-topic coverage.)
• The XP system tags effort. Showing every step earns more than guessing the right answer. No streak counter ships — for the longer reason, see why we don't use streaks.
• The focus timer surfaces a 2-minute warning and a 30-second warning before any chunk ends — visible on the timer the whole time. (Spoken transition cues — "two more problems, then a break" — are on the roadmap; the visual warnings ship today.)

And one thing about the bigger architecture: every camera frame, every word your child writes, every voice sample stays on the device in your home. Nothing leaves the device. We wrote about that decision here. For an ADHD kid whose data could otherwise become someone else's training corpus, that part matters too.

One more thing.

If the homework battle is bad enough that bedtime is on fire most nights, that's a sign — not of a bad kid, not of a bad parent — but that the load isn't matched to the kid. Talk to the teacher. Ask about an evaluation. Ask whether the assignment can be reduced. Most schools will work with you on this, especially if you have specific data ("Emma did six problems in 18 minutes; the assigned 20 would be 3 hours of work for her tonight"). It's not weakness to ask for an accommodation. It's the system working the way it's supposed to.

And on the days when none of this works and the homework just doesn't get done, that's also fine. One evening doesn't define a math student. The day-after-day pattern is what matters, and the pattern is shaped by the moves above more than by any one homework session.

If you want to know when Koda is available, drop your email here. We'll send a few real emails — not a newsletter, not a drip campaign.