AI Problem-Solving Framework for Middle School
Middle school is where the gap between routine and non-routine problem solving becomes critical. Students in grades 6-8 can typically execute known algorithms, but they struggle when problems require selecting a strategy, the exact skill tested on every standardized math assessment. The AI Problem-Solving Framework gives you guided questions for practising the metacognitive habits that distinguish students who can figure out what to do from students who can only do what they have been shown.
This page fills in a draft from a template. Copy prompt gives you the prompt for any AI chat.
- Polya's framework for grades 6-8
- 4 phases
- Pre-algebra, science, and design challenges
- 3 domains
- Metacognitive reflection prompts in every guide
- Look Back
How middle school teachers and students use the framework
Polya's 4 steps adapted for Middle School problem types.
Pre-algebra and early algebra word problems
Middle school algebra word problems require students to translate a real-world situation into an equation, a translation step that many students skip, going directly to computation and getting the wrong answer. The Understand step helps students identify the unknown before solving, and with Improve with AI on you can ask for prompts to name the variable and write a representation.
Rate, ratio, and proportion problems
Ratio and proportion problems are a middle school perennial stumbling block. The framework guides students to create a table of values (a strategy suggestion), see the proportional relationship, and then generalize, building the conceptual foundation that carries into slope and linear functions in high school.
Science data interpretation
In middle school science, students analyze data from experiments and must reason about what the data shows, a non-routine task that requires distinguishing correlation from causation, identifying outliers, and constructing evidence-based claims. The framework structures this reasoning with phase-specific questions for observation, hypothesis, and verification.
Engineering design challenges
Project-based design challenges (build a bridge with limited materials, design a water filter) are open-ended non-routine problems where Polya's phases apply directly: understand the constraints, devise a design plan, build and test, then look back and revise. The framework turns engineering challenges into structured problem-solving practice.
Test preparation for state assessments
State math assessments in grades 6-8 include multi-step extended response problems that are specifically designed to test non-routine problem solving. Practising the four-phase approach gives students a process to use on these items beyond algorithm recall.
Independent problem-solving practice
Teachers assign the framework for homework problems that are intended to develop thinking rather than practice a specific procedure. Students work through the four phases independently, writing brief responses at each step. This protocol makes mathematical thinking visible and reviewable, teachers can identify exactly where in the process a student's reasoning breaks down.
Looking for a different context? See all Problem-Solving Framework variants or browse all AI tools.
- No student accounts.
- No ads.
- Data-processing agreement on request.
What this page does here
This page builds a draft from a template on this device. Copy prompt gives you the prompt for any AI chat.
Questions teachers ask
Does the framework work for both pre-algebra and algebra students?
Yes. The guided questions adapt to the problem type the teacher types. Pre-algebra students working on ratio problems get questions appropriate for proportional reasoning. Students in algebra working on word-problem-to-equation translation get questions focused on identifying the unknown and writing a variable expression. The framework is problem-type-responsive, not curriculum-specific.
How does this differ from the math problem-solving steps my textbook already has?
Most textbook problem-solving steps are generic (Read, Plan, Solve, Check) and are listed without the specific guided questions that make each step actionable. The AI generates specific questions for each phase based on the actual problem the teacher types, making the process tailored rather than generic. The Look Back questions, in particular, are problem-specific and connect this solution to generalizable problem-solving habits.
Can the framework be used in science class as well as math?
Yes. The framework handles any non-routine problem where the approach needs to be figured out, not just the answer. Science problems involving experimental design, data interpretation, and evidence-based argumentation are excellent candidates. The Understand phase helps students distinguish what the evidence shows from what they want it to show, a critical scientific reasoning skill.
Should students use the framework on every problem or only difficult ones?
The framework is most valuable for non-routine problems, those where the student is not sure which approach to take. For routine practice problems (solving an equation, computing a percent), the framework adds unnecessary overhead. A useful rule of thumb: if a student knows which algorithm to apply within the first thirty seconds, it is a routine problem; if they do not, the framework is the right tool.
How do I assess whether students are actually using the four phases?
Ask students to submit a brief written record of their process alongside the solution, one sentence per phase is sufficient. This protocol makes thinking visible and allows you to identify at which phase the breakdown occurred for students who got the wrong answer. Over time, you will see students internalize the phases and apply them without the written scaffold.
Related tools
AI with your school's own student and course data is part of OpenEduCat.
Book a demoCounted per page. Report a problem for bugs.