My child can calculate but cannot solve word problems: what is missing?
By 15Q20M Editorial · Last updated
A child who computes correctly but keeps missing word problems usually has one of three gaps: misreading what's asked, picking the wrong operation from the words, or losing track mid-way through a multi-step problem. None of them are a math problem — they happen before any calculation starts. Here is how to tell which one, and what to practice.
Why calculating and solving word problems are different skills
A worksheet of “47 − 18” and a word problem that resolves to the same subtraction are not testing the same thing. The worksheet only tests calculation. The word problem also tests reading comprehension and translation — figuring out what is being asked and which operation answers it — before any calculation even starts. A child can be completely solid on the arithmetic and still miss the problem, because the part that failed was never the math.
In two large studies analyzing 6,595 errors made by 634 students on written word problems, about 40% of the errors happened before any calculation began — at the reading, comprehension, or operation-choice stage (Clements, 1980).
The 3 places word-problem solving actually breaks down
| Breakdown point | What it looks like | What it is not |
|---|---|---|
| Reading and understanding | Misses a key detail, skips a word that changes the problem, or cannot restate in their own words what is actually being asked | Not a comprehension-in-general problem — often only shows up under time pressure or with long problems |
| Choosing the operation | Understands the situation but picks the wrong operation — adds when the problem needs subtraction, or multiplies when it needs division | Not a computation error — the arithmetic that follows is usually done correctly for the (wrong) operation chosen |
| Tracking multiple steps | Gets the first step right, then loses track of what the first answer was supposed to feed into for step two or three | Not a sign they don't understand any of the individual steps — each one in isolation is usually fine |
A quick way to tell which one is the problem
- Cover the numbers and ask them to restate the question in their own words. If they can't, the gap is reading — not math at all.
- Ask what operation they'd use before they touch the numbers. A wrong answer here, with correct arithmetic afterward, points to operation choice.
- On a multi-step problem, ask what the answer to step one is going to be used for. If they can't say, they're losing the thread between steps, not making a calculation error.
- If none of these trip them up but the final answer is still wrong, look at the arithmetic itself — the gap may simply be computation, not the problem-solving layer.
How to help each gap without more worksheets
For reading and understanding, the fix is a restating habit, not more reading practice in general: before solving anything, say what the problem is asking in one sentence. This alone catches most misreads, because the habit forces a second look at the problem instead of the first number seen.
For operation choice, watch out for “keyword” shortcuts — “total means add,” “left means subtract.” They work on simple problems and actively mislead on multi-step or oddly worded ones. Ask what is actually happening in the situation instead: is something being combined, taken away, split evenly, or repeated? That question transfers to problems a keyword list won't cover.
For multi-step tracking, the fix is writing down what each number represents as they go — not just the number, but what it is. “12 (cookies left after the party)” carries into step two in a way a bare “12” does not. Losing track is usually a labeling problem, not a working-memory one.
All three benefit from the same underlying structure: a short, fixed daily set with feedback on what specifically went wrong, rather than a stack of mixed word problems and a single score at the end. That's the design behind our own daily practice — 15 problems, 20 minutes, one session a day on paper, with a topic breakdown and explanation on anything missed.
Try a free sample day — one full session, scored in the app
When word-problem trouble is bigger than word problems
Worth a conversation with the teacher if the same breakdown point keeps showing up across different topics (not just fractions, or just multiplication), if reading comprehension struggles show up outside of math too, or if a child who can explain the situation out loud still cannot translate it into an operation — that gap tends to need direct instruction, not just more practice.
If it's more than word problems, start here: my 5th grader is struggling with math
If wrong answers come from rushing rather than reading, see our 5-step error check
Fallen behind? A 4-week plan to help your child catch up in math
Ratio word problems: four questions that catch most 6th grade errors
Understands math at home but fails tests? How to tell a wording gap from a time gap
Frequently asked questions
Why can my child do math facts but not word problems?
Because word problems test an extra skill on top of computation: reading the situation and translating it into the right operation. Strong math facts don't help with that translation step — it's closer to a reading and reasoning skill than an arithmetic one, which is why fact fluency alone doesn't fix it.
Do keyword strategies (like “total means add”) help with word problems?
Only on the simplest problems, and they actively backfire on multi-step or unusually worded ones — a problem can include the word “total” and still require subtraction. Asking what's actually happening in the situation, rather than scanning for a trigger word, transfers to more problem types.
How many word problems should a child practice a day?
A handful done carefully, with the reasoning checked, teaches more than a full page done quickly. Two or three word problems where a child explains their operation choice out loud will do more for this specific skill than fifteen solved silently.
Sources
- White, A. L. (2009) — A Revaluation of Newman's Error Analysis (Mathematical Association of Victoria) — Newman (1977, 1983) developed a 5-level diagnostic (reading, comprehension, transformation, process skills, encoding) specifically for written mathematics word problems; Clements (1980) analyzed 6,595 errors by 634 children and found roughly 40% occurred before computation began. (verified 2026-08-26)
Research citations link to the original papers. Statistics are checked against their primary source.
See the routine in action.
15 problems, 20 minutes, scored in seconds — or compare us with other math workbooks.