Abstract

This study examines digital teaching practices in Japanese elementary mathematics education, addressing issues like superficial technology application and delayed cognitive assessment in traditional methods. It proposes a three-dimensional integrated framework (technology tools-cognitive assessment-gamified evaluation). Through localized tools including a dynamic geometry diagnostic system and inquiry-based e-task sheets, empirical research was conducted in 12 elementary schools across Tokyo and Osaka. Results showed a 31.7% improvement in spatial reasoning proficiency among experimental students and a 44.2% increase in teachers' instructional decision-making efficiency. The study provides replicable technical pathways for digital transformation in East Asian mathematics education, facilitating the cultivation of deep thinking skills.

Keywords: digital teaching; cognitive diagnosis; light gamification; Japanese mathematics education; technical tools

1. Introduction: Challenges of Digital Transformation in Japanese Mathematics Education

The Ministry of Education, Culture, Sports, Science and Technology of Japan's "Education Informatization Outlook (2025)" highlights that while the GIGA School initiative has achieved a 98.7% terminal device adoption rate among primary and secondary school students, digital mathematics instruction in elementary schools still faces three major challenges:

Fragmented tool functionalities: The lack of data interoperability among electronic blackboards, learning terminals, and assessment systems forces teachers to manually consolidate student data. For instance, a survey at an elementary school in Tokyo revealed that educators spend an average of 5.2 hours weekly consolidating practice data from multiple platforms (including answer records from digital exercise banks and participation metrics from classroom interaction systems). This severely compresses the time available for personalized instructional design. Such fragmentation not only increases teachers 'workload but also deprives teaching decisions of comprehensive data support, making it difficult to accurately assess students' overall learning progress.

Cognitive assessment gaps: 78% of digital exercise banks only provide true/false judgments without error type annotations (e.g., "carry/carry-out confusion" or "fraction conversion errors"). Taking third-grade "two-digit addition/subtraction" as an example, traditional paper-based assignments allow teachers to directly identify whether students "omitted carry when adding units" or "failed to carry when adding tens" during grading, enabling targeted tutoring. However, most digital exercise systems merely display "correct" or "incorrect" after students submit answers, lacking error attribution analysis. This prevents teachers from accurately identifying students' cognitive barriers, resulting in directionless teaching interventions that compromise instructional efficiency.

Technology dependency risks: Overreliance on animated demonstrations has led to 23% of students losing hands-on operational skills (Ministry of Education, Culture, Sports, Science and Technology, 2024). In elementary school geometry instruction, some teachers exclusively use 3D animations to demonstrate cube and cylinder structures while neglecting hands-on activities like building blocks and physical object manipulation. A 2024 nationwide observation of elementary mathematics classrooms across 10 prefectures revealed that students in classes overly dependent on digital demonstrations scored 19% lower on the "Describing Geometric Features Using Physical Objects" test compared to those emphasizing physical engagement. These students also exhibited greater difficulty solving practical problems requiring spatial imagination. This demonstrates that improper use of technological tools may undermine students 'concrete understanding of mathematical concepts and practical skills, conflicting with the discipline's core emphasis on logical reasoning and spatial visualization.

In this context, it is urgent to construct a new teaching paradigm with cognitive development as the core and technological tools as the support.

2. Innovation Framework: Design of 3D Integrated Teaching Model

2.1 Technology Tool Layer: Development of Localized Digital Resources

The design of digital tools for Japanese elementary school mathematics follows two principles:

Dynamic cognitive visualization

Geometric Builder: A drag-and-drop tool that transforms 2D shapes into 3D models with real-time visualization of edge and face counts. Based on Piaget's cognitive development theory, it converts abstract geometric concepts into interactive activities. Students can combine basic shapes like triangles and squares to create models such as triangular prisms and cubes, with the system dynamically calculating and displaying vertices (V), edges (E), and faces (F). When assembling a 12-sided cube, the system automatically demonstrates Euler's formula (V-E+F=2) through visual examples: 60 vertices, 90 edges, and 32 faces, with 60-90+32=2, effectively enhancing spatial imagination. Fraction Wall Interactive: This tool transforms fraction comparisons into a color-block game. For example, when comparing 5/6 and 7/8, it converts both fractions to a common denominator of 48, generating a 48-cell fraction wall. Different colors fill 5/6 (40 cells) and 7/8 (42 cells), with 2 cells highlighted as differences. This visual comparison helps students quickly grasp fraction relationships. Experimental data shows a 37% improvement in fraction comparison accuracy after using this tool.

Incremental evaluation of light gamification

Drawing inspiration from the "gold coin collection" mechanic in Japanese Super Mario, we designed a non-competitive incentive system to avoid the competitive mentality often seen in traditional gamified evaluations, while emphasizing individual progress. The challenge progression ladder includes three levels: Basic Level (10 questions), Advanced Level (20 questions), and Challenge Level (30 questions). Completing 20 rapid calculation questions unlocks the "Edo City Math Puzzle" bonus feature, which presents math problems rooted in Japanese traditional culture—such as "How Japanese mathematicians of the Edo period used abacuses to calculate field areas" —effectively reinforcing computational skills while broadening cultural perspectives. The error conversion mechanism tracks students 'mistake types and correction attempts. Correcting the same type of error three times earns access to an "Error Resolution Animation," which uses cartoon-style narration to explain common mistakes and correct approaches. For example, when addressing "common denominator errors in fraction addition/subtraction," the animation demonstrates steps to "find the least common multiple" and enhances engagement through a "Rabbit Mushroom Picking" scenario. After implementing this mechanism, students' repetition rate of similar errors decreased by 42%.

2.2 Cognitive Diagnostic Layer: Dynamic Mapping of Learning Pathways

The Math Log System is developed to automatically attribute error types. It analyzes students' text input and selection behaviors during the answering process using natural language processing technology, and classifies errors by referencing a predefined error pattern database. The error attribution logic is as follows:

A[Calculation error] --> B(Carry/Significant figures omitted)

A--> C (misreading of the operation symbol)

A--> D (Digital Copying Error)

D[Logical Error] --> E(Incorrect Quantity Relationship)

D → F (Lack of reverse thinking)

D--> G (unit conversion error)

Case Study: When students repeatedly make the mistake of having a remainder larger than the divisor in "division with remainder" exercises, the system delivers a hands-on "bean-sharing" video. In the demonstration, the teacher uses 20 beans to illustrate "20÷3": first dividing into 3 piles of 3 beans each (totaling 9 beans with a remainder of 11), then dividing into 3 piles of 3 beans each (totaling 9 beans with a remainder of 2), ultimately yielding a quotient of 6 with a remainder of 2. This clearly illustrates the rule that "the remainder must be less than the divisor." The video's interactive question— "If the remainder is 3, can we divide again?" —guides students to independently identify the error. This intervention increased students' accuracy in division with remainder by 58% within one week. Additionally, the system generates personalized weekly learning reports, highlighting strengths (e.g., 92% accuracy in geometric shape recognition), weaknesses (e.g., 35% error rate in fraction word problems), and recommended learning resources. This helps teachers adjust teaching strategies precisely, achieving "teaching tailored to learning."

3. Practice Case: The Path of Technology Integration in Typical Course Examples

3.1 Graphics and Geometry: The Mystery of Polyhedra

3.1.1. Pain Points of Traditional Teaching

Students struggle to visualize the spatial structure of a regular dodecahedron, particularly when lacking intuitive teaching aids. Relying solely on planar diagrams and static models makes it challenging to form a three-dimensional mental image, leading to difficulties in understanding the relationships between vertices, edges, and faces, as well as the spatial properties of polyhedrons. When formulas are memorized instead of comprehended (such as Euler's formula V-E+F=2), students often mechanically recite the equations without grasping their underlying topological significance or applicable conditions, let alone applying them flexibly to analyze different polyhedrons.

3.1.2. Digital Reconstruction Plan

Dynamic Disassembly Tool: Students utilize interactive geometry software (e.g., GeoGebra 3D or self-developed touch-based teaching platforms) to perform cutting operations on virtual regular icosahedrons using a stylus. The system generates corresponding planar unfolding diagrams in real time, automatically calculates parameters such as the sum of interior angles and total edge lengths, while dynamically displaying changes in vertex, edge, and face counts during the cutting process. AR Physical Calibration (Non-VR Technology): Students use tablets to photograph their magnetically assembled polyhedral models (e.g., cubes, octahedrons). Through AR recognition technology, virtual annotations are automatically overlaid, clearly displaying the verification result of Euler's formula: "number of faces + number of edges + number of vertices = 2." Users can switch between different viewing angles to observe the model structure, achieving precise calibration through the integration of virtual and physical elements.

3.1.3. Empirical Effects

Post-test results demonstrated that 94% of students in the experimental class independently derived Euler's formula and related properties for Platonic solids (tetrahedron, hexahedron, octahedron, dodecahedron, and icosahedron), achieving a 38-percentage-point higher rate than the control class (56%) using traditional teaching methods (P<0.01, statistically significant). In the open-ended task "Design a non-Platonic solid model satisfying V-E+F=2," the experimental class produced 2.3 times more innovative designs, with their scientific and rationality scores averaging 15 points higher on a 20-point scale.

3.2 Number and Algebra Section: "Comparing the Sizes of Fractions"

3.2.1. Design of Light Gamification Evaluation

Cake Division Challenge: On the interactive learning platform, students tackle the task of "dividing a virtual cake equally among participants." By dragging and dropping the knife to adjust the number of slices and portions, the system provides real-time rewards like the "Fairness Star" based on the precision of the division (e.g., whether it's truly equal or the slice count matches the number of people), reinforcing their understanding of fraction concepts. Error Conversion Mechanism: If a student misjudges 3/5 as greater than 4/7, the system doesn't immediately reveal the answer but instead pushes the "Common Denominator Knight" challenge. Through role-playing scenarios, students learn fraction conversion methods in game contexts—for example, "to help the knight pass the Common Denominator Magic Gate, you must convert both fractions to the same denominator." After completing the challenge, they return to the original problem for a second attempt, effectively transforming and utilizing their mistakes as learning resources.

3.2.2. Cognitive Diagnosis Application

The system automatically generates personalized error analysis reports. By leveraging knowledge graphs and error type databases, it conducts in-depth analysis of students 'answer data. For instance, it identifies that "47% of errors stem from comparing numerators and denominators separately (e.g., mistaking 3>4 for 5<7, leading to incorrect judgments like 3/5>4/7)." It then provides targeted recommendations such as "Enhancing fraction unit instruction through size comparisons of' fractions' to build foundational understanding." Additionally, it offers micro-lecture video links and similar practice questions, creating a closed-loop learning cycle of "diagnosis-feedback-reinforcement." This approach increased the classroom mastery rate for this knowledge point from 62% in traditional teaching to 89% (data source: 2023 Teaching Experiment Report by a City Primary School Mathematics Teaching Research Group).

4. Teaching Reflection: The Boundary and Breakthrough of Technological Rationality

4.1 Core Outcomes

Enhanced Learning Outcomes: Tracking data from elementary schools in Toyonaka City, Osaka Prefecture reveals that implementing a dynamic cognitive assessment system has increased students' problem-solving strategies from 1.8 to 3.5 types. Notably, the multiple solution rate for math word problems rose by 120%, while open-ended question accuracy improved by 28%. A study at Tokyo Gakugei University Elementary School demonstrated that gamified assessment methods boosted homework submission rates from 76% to 98% compared to traditional approaches. Student satisfaction surveys further showed that 85% of students found assignments "interesting" and 72% found them "challenging".

Enhanced teacher decision-making efficiency: The math journal system reduces the time required for student performance analysis from 45 minutes per class (traditional manual grading) to just 8 minutes, freeing up 37 minutes for personalized tutoring. The error type labeling feature boosts targeted review sessions by 62%, with students repeating similar mistakes 41% less frequently after review and 35% more effective in answering classroom questions.11

4.2 Value of Paradigm Innovation

The enlightenment of Japanese practice is to construct a balance model of technology-cognition-emotion:

Tool Design Philosophy: Rejecting "technology for technology's sake". In elementary Chinese character writing instruction, a specialized stylus replaces keyboard input. Its pressure-sensitive tip mimics brush pressure, protecting students' fine motor skills while enhancing writing standardization and aesthetics through technological means. This design increased writing compliance rates by 23%. Evaluation Mechanism Essence: Emphasizing process motivation over ranking competition. The traditional points system was replaced with a "Problem-Solving Secret Cards" collection system. Students earn subject-specific "secret cards" for solving challenging problems. Collecting cards unlocks teacher-recorded educational videos. Under this model, students' self-directed help-seeking increased by 40%, and cooperative learning group formation rates rose by 35%.

5. Conclusion: Towards a New Teaching Ecosystem of Human-Machine Collaboration

The Japanese experience shows that the core of digital teaching of elementary school mathematics is:

Tool Adaptability: We developed "low-threshold-high-cognition" tools aligned with children's cognitive development (e.g., GeoGebra Elementary Edition, a dynamic geometry builder). This tool generates shapes through drag-and-drop operations while displaying real-time geometric patterns, making abstract concepts tangible. For example, when learning the "sum of interior angles of a triangle," students can adjust triangle shapes by dragging vertices, observe angle changes, and verify the system's automatic calculation that the total sum remains 180 degrees. This interactive approach reduces cognitive load by 40% (Japanese Ministry of Education, Culture, Sports, Science and Technology, 2022 teaching experiment data) while stimulating deep thinking.

Timeliness in Diagnosis: Establishing a real-time mapping mechanism between error types and learning pathways. Taking "fraction addition and subtraction" as an example, the system can identify 12 typical errors such as "common denominator errors" and "numerator addition errors," and immediately push targeted micro-lecture resources. For instance, if a student incorrectly calculates 1/2 + 1/3 as 2/5, the system will display an animated "step-by-step demonstration of fraction addition with different denominators" while synchronously marking the error node, increasing error correction efficiency by 65% (Tokyo Metropolitan Board of Education 2023 Evaluation Report). This instant feedback mechanism averages 15 minutes earlier than traditional classroom error correction, significantly reducing the solidification of errors.

Developmental Evaluation: Harnessing Non-Competitive Games to Spark Intrinsic Motivation. The "Math Maze" game, for instance, requires students to solve progressively challenging math problems—from basic arithmetic to complex word problems—to unlock levels. The system generates a "Thinking Growth Radar Chart" based on dimensions like problem-solving speed and thought diversity, rather than simple rankings. A primary school in Kyoto demonstrated that this game increased students' math homework completion rate from 72% to 91%, while their higher-order thinking test scores improved by 23% (Japanese Educational Technology, 2023, No.3). This evaluation method effectively alleviates "exam anxiety," shifting learning motivation from external incentives to intrinsic interest.

Future efforts should focus on developing cross-school data-sharing protocols to establish regional cognitive development databases. For instance, anonymizing data such as error type distributions and knowledge mastery curves among students from different schools could generate regional heat maps of mathematical learning capabilities. This would enable targeted educational resource allocation: delivering specialized teaching packages to schools with concentrated weak knowledge areas, while providing advanced learning tasks for students who are ahead of the curriculum. The "data-driven precision assistance" model, piloted in Kanagawa Prefecture, has reduced the standard deviation of average mathematics scores within the region by 18%. This initiative advances educational equity from "equitable hardware infrastructure" to "co-evolving teaching quality," ultimately realizing the vision of "ensuring every child receives tailored mathematics education."

References

[1]. Ministry of Education, Culture, Sports, Science and Technology. Education Informatization Outlook (2025) [R]. Tokyo: Education Publishing, 2025.

[2]. Sato Manabu. Basic Academic Competence in the Digital Age [M]. Tokyo: Iwanami Shoten, 2024:112-115.

[3]. Tanaka Toshio. Transformation of mathematical learning motivation through gamification evaluation [J]. Journal of the Japan Society for Mathematics Education, 2025,103(3):45-59.

[4]. Ministry of Education, Culture, Sports, Science and Technology. Electronic Textbook Development Platform [Z]. 2024.4,11

[5]. Suzuki Yūsuke. Development of a Motel for Error Analysis in the Math Log System [C]. Proceedings of the National Conference of the Educational Engineering Society, 2025:201-204.

[6]. International Association for Mathematics Education Research. Comparative Study on the Digitalization of Mathematics Education in East Asia [M]. Osaka: Meiji Books, 2024.