---
license: cc-by-4.0
task_categories:
  - visual-question-answering
  - image-classification
  - text-generation
language:
  - zh
tags:
  - education
  - math
  - error-analysis
  - handwritten
  - multimodal
  - scratchwork
pretty_name: ScratchMath
size_categories:
  - 1K<n<10K
configs:
  - config_name: primary
    data_files: "primary/data-*.parquet"
  - config_name: middle
    data_files: "middle/data-*.parquet"
dataset_info:
  - config_name: primary
    features:
      - name: question_id
        dtype: string
      - name: question
        dtype: string
      - name: answer
        dtype: string
      - name: solution
        dtype: string
      - name: student_answer
        dtype: string
      - name: student_scratchwork
        dtype: image
      - name: error_category
        dtype:
          class_label:
            names:
              0: 计算错误
              1: 题目理解错误
              2: 知识点错误
              3: 答题技巧错误
              4: 手写誊抄错误
              5: 逻辑推理错误
              6: 注意力与细节错误
      - name: error_explanation
        dtype: string
    splits:
      - name: train
        num_examples: 1479
  - config_name: middle
    features:
      - name: question_id
        dtype: string
      - name: question
        dtype: string
      - name: answer
        dtype: string
      - name: solution
        dtype: string
      - name: student_answer
        dtype: string
      - name: student_scratchwork
        dtype: image
      - name: error_category
        dtype:
          class_label:
            names:
              0: 计算错误
              1: 题目理解错误
              2: 知识点错误
              3: 答题技巧错误
              4: 手写誊抄错误
              5: 逻辑推理错误
              6: 注意力与细节错误
      - name: error_explanation
        dtype: string
    splits:
      - name: train
        num_examples: 241
---

# ScratchMath

**ScratchMath** is a multimodal benchmark dataset for evaluating the ability of Multimodal Large Language Models (MLLMs) to analyze handwritten mathematical scratchwork produced by real students. Unlike existing math benchmarks that focus on problem-solving accuracy, ScratchMath targets **error diagnosis** — identifying what type of mistake a student made and explaining why.

## Dataset Description

The dataset contains 1,720 samples of authentic student scratchwork collected from an online education platform. Each sample includes:

- A math problem with its correct answer and solution
- The student's incorrect answer
- A photograph of the student's handwritten scratchwork
- A human-annotated error category and detailed error explanation

### Subsets

| Subset | Description | Samples |
|--------|-------------|---------|
| `primary` | Primary school (Grades 1–6) math problems | 1,479 |
| `middle` | Middle school (Grades 7–9) math problems | 241 |

### Error Categories

Each sample is labeled with one of seven error categories:

| Category | English | Primary | Middle |
|----------|---------|---------|--------|
| 计算错误 | Calculation Error | 453 | 113 |
| 题目理解错误 | Question Comprehension Error | 499 | 20 |
| 知识点错误 | Knowledge Gap Error | 174 | 45 |
| 答题技巧错误 | Problem-Solving Strategy Error | 118 | 17 |
| 手写誊抄错误 | Handwriting Transcription Error | 95 | 29 |
| 逻辑推理错误 | Logical Reasoning Error | 73 | 2 |
| 注意力与细节错误 | Attention & Detail Error | 67 | 15 |

## Fields

| Field | Type | Description |
|-------|------|-------------|
| `question_id` | string | Unique identifier for the math problem |
| `question` | string | The math problem text (may contain LaTeX in `$...$` delimiters) |
| `answer` | string | The correct answer(s) |
| `solution` | string | Step-by-step solution with explanation |
| `student_answer` | string | The student's incorrect answer |
| `student_scratchwork` | image | Photograph of the student's handwritten work |
| `error_category` | ClassLabel | One of 7 error types (see above) |
| `error_explanation` | string | Detailed explanation of what error the student made and why |

## Usage

```python
from datasets import load_dataset

# Load primary school subset
ds_primary = load_dataset("songdj/ScratchMath", "primary")

# Load middle school subset
ds_middle = load_dataset("songdj/ScratchMath", "middle")

# Access a sample
sample = ds_primary["train"][0]
print(sample["question"])
print(sample["error_category"])
sample["student_scratchwork"].show()  # Display the scratchwork image
```

## Citation

If you use this dataset, please cite:

```bibtex
@inproceedings{song2026scratchmath,
  title={Can MLLMs Read Students' Minds? Unpacking Multimodal Error Analysis in Handwritten Math},
  author={Song, Dingjie and Xu, Tianlong and Zhang, Yi-Fan and Li, Hang and Yan, Zhiling and Fan, Xing and Li, Haoyang and Sun, Lichao and Wen, Qingsong},
  booktitle={Proceedings of the 27th International Conference on Artificial Intelligence in Education (AIED)},
  year={2026}
}
```

## License

This dataset is released under the [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) license.