How to Write Math Problem-Solving IEP Goals

Quick Answer

A strong math problem-solving goal measures how the student represents and solves defined word problems, not just raw computation. Use comparable one- or multi-step probes, record accuracy and prompt level, and separate setup/operation-choice errors from calculation errors so the data show which part of problem solving is changing.

For math problem solving, the fastest way to make this usable is to decide what staff will actually count. For math problem solving, the core measurement is percent correct on defined one- or multi-step word problems, with prompt level and strategy-use data when independence matters.

What a strong math problem solving goal includes

A problem-solving goal needs more than a percent correct. Use Condition → Behavior → Criterion to identify the problem type and number of steps, the strategy or organizer available, and whether calculation tools are permitted. The behavior should capture solving the word problem or completing defined solution steps. The criterion can include final-answer accuracy plus prompt level or strategy independence when those are part of the student's actual need.

Measurability comes from 34 CFR §300.320(a)(2); the regulation does not define a federal percent-correct benchmark for word problems. Use comparable problem types and specify prompting, organizer use, or calculator access when those conditions affect the student's baseline.

Problem representation — % problems set up correctly: Shows whether the student translates language into a mathematical model. Solution accuracy — % complete problems correct: Useful when reasoning and computation are both in scope. Independence — prompt code per problem: Shows whether success depends on adult cueing.

Describe the word-problem probe so another teacher can build a comparable set: operation, number of steps, reading load, organizer or calculator access, and scoring for partial solution steps. Separate reasoning from calculation. A student may select the correct operation and still miss the final answer because of a computation error, which one total percentage can hide.

Problem solving measures selecting and applying a mathematical plan in context. Calculation measures executing operations; the two can be related without being the same goal.

When you need the next layer of the workflow, use browse a full math goal bank, math calculation goals, Condition–Behavior–Criterion formula.

Use error coding to protect the reasoning signal

When a word-problem answer is wrong, mark where the solution failed. A compact code can separate problem representation, operation selection, computation, multi-step sequencing, and answer labeling. That matters because two students with 50% final-answer accuracy may need completely different instruction: one chooses the wrong operation; the other chooses correctly and then makes regrouping errors.

If calculator use is an accommodation or an instructional support, decide whether it belongs in the goal condition. A calculator can reduce computation load so the team can measure reasoning more cleanly. If the student's need is independent calculation, the opposite may be true. The measurement choice should follow the construct the team wants to monitor, not a blanket rule that calculators make a problem-solving probe easier or less valid.

How to get a baseline

Use several comparable word-problem samples to estimate current problem-solving performance; one unusually difficult worksheet is not a defensible baseline. Useful sources include:

Three parallel sets of word problems matched by operation and number of steps. Work-sample analysis that separates setup, operation choice, computation, and answer labeling. Structured probes that record the highest prompt needed.

Give two or three matched sets of word problems that use the same operations and number of steps. Score more than the final answer: note whether the student identified relevant information, chose an operation, set up the problem, computed accurately, and labeled the answer. Record the highest prompt used so a correct solution after extensive adult support is not treated as independent baseline performance.

Record problem type, number of steps, reading-access condition, allowed representations, prompt level, and final accuracy. When possible, code where the solution broke down—problem representation, operation/plan selection, or computation—so the team does not write a reasoning goal from a computation-only problem.

Separate reasoning from calculation. A student can choose the correct operation and then make a computation error; one total score can hide that distinction.

Example math problem solving goals

These word-problem goals are models of measurable structure. Replace the operation types, number of steps, supports, prompt limits, baseline, and target with values that match the student's own problem-solving data. Individualize the annual goal from those data before it goes into the IEP.

K–2 — Given a one-step addition or subtraction word problem within 100 and access to a visual organizer, the student will identify the operation and solve correctly in 4 of 5 problems across 3 probes. (Condition: one-step word problems and a visual organizer | Behavior: select the operation and solve | Criterion: 4 of 5 across 3 probes) How to measure: 5-item parallel problem-solving probes.

3–5 — Given multi-step word problems using whole numbers, the student will identify relevant information, select operations, and solve with at least 80% accuracy across 3 probes with no more than one verbal prompt. (Condition: multi-step whole-number word problems | Behavior: complete the problem-solving steps | Criterion: 80% across 3 probes with ≤1 verbal prompt) How to measure: task-analysis score plus final-answer accuracy.

3–5 — Given a word problem involving fractions, the student will represent the problem with an equation or model and solve correctly in 4 of 5 trials across 3 sessions. (Condition: fraction word problems | Behavior: represent and solve the problem | Criterion: 4 of 5 trials across 3 sessions) How to measure: score representation and solution on matched probes.

6–8 — Given a two-step ratio, percent, or proportional-reasoning problem, the student will choose a strategy and solve with at least 80% accuracy across 4 probes independently. (Condition: two-step ratio or percent problems | Behavior: choose a strategy and solve | Criterion: 80% across 4 probes independently) How to measure: curriculum-based 5-item probes.

6–8 — Given a real-world algebraic word problem, the student will define a variable, write an equation, and solve it correctly in 4 of 5 opportunities across 3 probes. (Condition: an algebraic word problem | Behavior: model with an equation and solve | Criterion: 4 of 5 across 3 probes) How to measure: 3-step task-analysis rubric.

9–12 — Given a multi-step applied math problem from the student’s course, the student will identify givens, select a method, show work, and reach a reasonable solution with 80% rubric accuracy across 3 samples. (Condition: a course-aligned applied problem | Behavior: use a complete problem-solving process | Criterion: 80% rubric accuracy across 3 samples) How to measure: score a defined problem-solving rubric, not computation speed.

Separate representation from computation in the data sheet

For multistep word problems, one total percent-correct score can hide the point of breakdown. Consider scoring three components during baseline: represents the problem, selects an appropriate operation or plan, and carries out the computation. A student who chooses the right operations but makes arithmetic errors needs a different instructional response from a student who cannot identify what the problem is asking.

You do not have to turn all three components into one annual goal. Use the analysis to choose the skill that most limits access. If computation is already addressed elsewhere, the problem-solving goal can focus on representation and strategy selection while staff continue to record whether arithmetic errors affected the final answer.

How to write the matching present level

The PLAAFP should report current success on defined word-problem types and identify the component that most limits independent problem solving. That educational-effect sentence should explain what the breakdown does to grade-level participation, not merely restate the percentage.

Worked present-level example: On three sets of five two-step word problems, Devon selected appropriate operations on 11/15 problems but completed only 7/15 accurately; four errors occurred after a correct setup. He needs support maintaining the solution sequence and checking computation.

Keep problem complexity and reading supports stable enough for trend interpretation. If the student begins solving more complex structures, mark the change and avoid comparing the new series as though the tasks were identical.

Progress monitoring

Schedule comparable problem-solving probes before the first reporting window begins. Probe every 2–3 weeks with parallel problem sets. Record setup accuracy and final-answer accuracy separately; add prompt level if adult support is part of the baseline. If final accuracy stays flat while setup accuracy rises, do not assume problem solving is unchanged. The data may point to a calculation bottleneck that needs instruction or a separate calculation goal.

Common mistakes

Before the goal is finalized, score a small set of representative word problems with the proposed rule.

Counting only the final answer and ignoring correct reasoning. Mixing one-step and multi-step items unpredictably. Changing calculator access between probes. Writing a computation goal under a problem-solving label. Using vague verbs such as 'understand' or 'improve'. Failing to define what counts as an independent response.

Score one sample word-problem set with another educator. If you disagree about partial credit, prompting, or what counts as a correct setup, resolve those rules now rather than after the first progress report.

FAQ

How is a math problem-solving goal different from a calculation goal?

Problem solving measures selecting operations, representing information, planning steps, and solving a contextual task. Calculation measures carrying out numeric procedures accurately or fluently. A student can calculate correctly but choose the wrong operation in a word problem. Separate measures help the team identify whether the barrier is reasoning, computation, language, or a combination.

Should I score only the final answer on multi-step word problems?

No, not when the instructional target includes the process. A final-answer score can hide whether the student selected the correct operation, organized the information, completed intermediate steps, or made one arithmetic error at the end. Use a step rubric or error codes when those distinctions matter, while still reporting overall accuracy if it is useful.

How do prompts fit into a math problem-solving criterion?

Define prompt level as part of the condition or criterion instead of writing only 'with support.' For example, track independence, one verbal cue, a visual organizer, or multiple adult prompts. A student may solve more problems correctly while remaining dependent on prompts, so accuracy and independence often need to be recorded as separate data fields.

What should a baseline probe for word-problem solving include?

Use several problems that represent the structures the student is expected to solve, such as one-step, multi-step, comparison, or ratio situations. Record accuracy and the error pattern, not just the total. Three comparable probe points can show whether the student’s difficulty is stable and whether one problem type is disproportionately difficult.

Can I use classroom assignments to monitor problem-solving progress?

Yes, if the assignments are comparable enough to support a fair trend. Permanent products are useful, but changing unit content, teacher help, calculator access, or problem complexity can make percentages misleading. For cleaner progress data, pair classroom work with a brief standardized classroom probe or a consistent rubric scored under documented conditions.

Need goals customized to your student?

Paste your assessment data and generate an IDEA-compliant, individualized goal in seconds.

Try it free →