Stepwise helps teachers find exactly where students struggle in math problems
STEPwise: Supporting Teachers in Diagnosing Students' Mathematical Problem-Solving Skills through LLM-Generated Questions and Rubric
Human-Computer Interaction
Summary
Solving math problems is tricky, and teachers often find it hard to figure out exactly where a student gets stuck. The authors developed STEP, a way to break down problem-solving into stages and skills, which helps identify specific points where students have trouble. They also built STEPwise, a tool that uses artificial intelligence to create tailored questions and scoring guides so teachers can check students’ thinking more easily. Tested in classrooms, this system helped both teachers and students understand problem-solving obstacles more clearly.
What this means in practice
- •For edtech developers: Create AI-powered tools that automatically generate stage-specific math diagnostic questions and rubrics tailored to classroom needs.$Commercial implications: Enables educational software companies to offer scalable diagnostic support tools that reduce teacher workload and improve math skill assessment.
- •For curriculum designers: Incorporate STEP-based diagnostic frameworks to design teaching materials that target specific problem-solving skills and reasoning stages.
Authors
Yoonsu Kim, Hayeon Doh, Eunhye Kim, Jinho Son, Juho Kim
Abstract
Problem-solving is central to mathematics education. However, diagnosing where students' reasoning breaks down requires individualized probing and consistent criteria that teachers cannot sustain at classroom scale. We introduce STEP (STage-based Evaluation of Problem-solving skills), co-designed with six mathematics education experts, which operationalizes problem-solving as four stages and eight diagnostic sub-skills. By specifying what evidence to elicit and assess at each stage, STEP enables an LLM pipeline to generate stage-specific questions and rubrics for any math word problem, reducing the manual design of diagnostic materials. An expert review (N=8) assessed the pedagogical validity of the generated materials and informed STEPwise, a teacher-in-the-loop system for diagnostic material generation and process-level assessment. In a classroom study with 5 teachers and 64 students, STEPwise helped teachers and students locate reasoning breakdowns and supported students in reflecting on their reasoning. Together, STEP and STEPwise offer a structured approach to fine-grained diagnosis at classroom scale.