Ben-45/nigerian_education_continuous_assessment

Nigeria Education – Continuous Assessment

Dataset Description

Mid-term and end-of-term CA scores with attendance correlation.

Category: Student Performance & Exams
Rows: 600,000
Format: CSV, Parquet
License: MIT
Synthetic: Yes (generated using reference data from WAEC, JAMB, UBEC, NBS, UNESCO)

Dataset Structure

Schema

• student_id: string
• subject: string
• term: string
• year: integer
• attendance_rate: float
• ca_score: float

Sample Data

| student_id   | subject     | term   |   year |   attendance_rate |   ca_score |
|:-------------|:------------|:-------|-------:|------------------:|-----------:|
| STU-0288036  | Mathematics | Term 3 |   2024 |              0.7  |       59.1 |
| STU-0099093  | Physics     | Term 3 |   2024 |              0.91 |       90.5 |
| STU-0035019  | Economics   | Term 3 |   2024 |              0.52 |       97.6 |
| STU-0204246  | Physics     | Term 2 |   2022 |              0.83 |       63.9 |
| STU-0478928  | Mathematics | Term 1 |   2024 |              0.67 |       67.5 |

Data Generation Methodology

This dataset was synthetically generated using:

1. Reference Sources:

   • WAEC (West African Examinations Council) - exam results, pass rates, grade distributions
   • JAMB (Joint Admissions and Matriculation Board) - UTME scores, subject combinations
   • UBEC (Universal Basic Education Commission) - enrollment, infrastructure, teacher data
   • NBS (National Bureau of Statistics) - education surveys, literacy rates
   • UNESCO - Nigeria education statistics, enrollment ratios
   • UNICEF - Out-of-school children, gender parity indices

2. Domain Constraints:

   • WAEC grading system (A1-F9) with official score ranges
   • JAMB UTME scoring (0-400 points, 4 subjects)
   • Nigerian curriculum structure (Primary, JSS, SSS)
   • Academic calendar (3 terms: Sep-Dec, Jan-Apr, May-Jul)
   • Regional disparities (North-South education gap)
   • Gender parity indices by region and level

3. Quality Assurance:

   • Distribution testing (WAEC grade distributions match national patterns)
   • Correlation validation (attendance-performance, teacher quality-outcomes)
   • Causal consistency (educational outcome models)
   • Multi-scale coherence (student → school → state aggregations)
   • Ethical considerations (representative, unbiased, privacy-preserving)

See QUALITY_ASSURANCE.md in the repository for full methodology.

Use Cases

• Machine Learning: Performance prediction, dropout forecasting, admission modeling, resource allocation
• Policy Analysis: Education program evaluation, gender parity assessment, regional disparity studies
• Research: Teacher effectiveness, infrastructure impact, exam performance patterns
• Education Planning: School placement, teacher deployment, budget allocation

Limitations

• Synthetic data: While grounded in real distributions from WAEC/JAMB/UBEC, individual records are not real observations
• Simplified dynamics: Some complex interactions (e.g., peer effects, teacher-student matching) are simplified
• Temporal scope: Covers 2022-2025; may not reflect longer-term trends or future policy changes
• Spatial resolution: State/LGA level; does not capture micro-level heterogeneity within localities

Citation

If you use this dataset, please cite:

@dataset{nigeria_education_2025,
  title = {Nigeria Education – Continuous Assessment},
  author = {Electric Sheep Africa},
  year = {2025},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/electricsheepafrica/nigerian_education_continuous_assessment}
}

Related Datasets

This dataset is part of the Nigeria Education Sector collection:

• https://huggingface.co/collections/electricsheepafrica/nigeria-education-sector

Contact

For questions, feedback, or collaboration:

• Organization: Electric Sheep Africa
• Collection: Nigeria Education Sector
• Repository: https://github.com/electricsheepafrica/nigerian-datasets

Changelog

Version 1.0.0 (October 2025)

• Initial release
• 600,000 synthetic records
• Quality-assured using WAEC/JAMB/UBEC/NBS reference data