The hardest part of being a math parent is knowing when to intervene. Hover too close and you end up doing the homework yourself, which teaches the child that struggle is bad and adults will rescue them. Hover too far and they give up and hate math. The right altitude is tricky and it changes from problem to problem.

Here is a three-move heuristic that works for most elementary and middle-school math homework. When your child is stuck, start with move one and only escalate if the previous move didn't unstick them.

Move one — ask what they know. "What does the problem want you to find?" "What numbers did it give you?" These are low-stakes, retrieval-only questions that re-engage the child's working memory. Seventy percent of the time this alone unsticks them.

Move two — ask what they've tried. If the child has tried something and it didn't work, ask them to walk you through it. Often, hearing themselves out loud is enough for them to spot the error. If not, you can ask an *open* question about their attempt: "what made you choose addition there?" Do not correct them yet.

Move three — model one similar problem, then step away. If moves one and two didn't unstick them, solve a *similar but not identical* problem out loud while they watch. Think aloud — say the steps you're considering. Then leave them to attempt the original problem themselves. Do not sit next to them while they do it.

What this heuristic avoids is the failure mode of "explaining the problem until the child nods". Explaining is very efficient for *you*, the adult who already knows the answer, and very inefficient for the child who needs to *construct* the understanding themselves. If your child is nodding a lot and still getting the problem wrong, you are probably explaining too much.

One final piece of advice. Frustration is a signal, not a catastrophe. A child who says "I can't do this!" and throws a pencil is not broken. They are experiencing a normal cognitive event. The best response is a short break — five minutes is often enough — followed by a return to the problem with fresh eyes. Returning is the key; abandoning the problem reinforces that frustration means give up, which is the lesson we really don't want them to learn.

One small adjustment that helps almost every family: separate the sit-down from the start. A child who is dreading homework will resist starting the work, and ten minutes of arguing about whether they have to start sets a sour tone before any pencil has moved. If you can make the sit-down a routine — same time, same chair, same opening question — the resistance dissolves into "this is just what we do at four o'clock" and the actual work begins faster. The routine takes a couple of weeks to settle and then it pays you back forever.

Another useful move is the "two-minute warm-up." Before the homework itself, ask the child to do a single very-easy problem — a fact they already know cold, an addition that's well within range. They get it right, the brain registers a small win, the internal monologue shifts from "this is going to be terrible" to "I can do this." Then the actual homework feels like the next problem in a sequence rather than a fresh confrontation. The warm-up is not the homework. It's the on-ramp.

When you do help, help on the process, not on the answer. "Try it without me first; I'll come back in three minutes" is one of the most powerful sentences in the parental homework-help vocabulary. So is "show me what you've tried so far." So is "what is this question really asking?" These are all things a teacher would say, and they invite the child to do the actual cognitive work — which is, after all, the entire point of homework existing. If you give the answer, the homework is done. If you give the prompt, the child does the homework. Only one of those changes anything.