Technical Report TR-004 — Validation Protocol
Describe the scientific protocol used to validate the outputs generated by SportsLabResearch-WIMU-DataExtractor.
Purpose
Define the methodology used to assess the accuracy, consistency, and reproducibility of the extracted datasets.
Validation Strategy
Validation compares the software outputs with the corresponding reference datasets exported from WIMU® software.
Validation Workflow
Script Output
│
▼
Reference Dataset
│
▼
Variable Matching
│
▼
Statistical Analysis
│
▼
Validation Report
Validation Metrics
| Metric | Description |
|---|---|
| Mean Difference (MD) | Average difference between methods. |
| Mean Absolute Error (MAE) | Mean absolute error. |
| Root Mean Square Error (RMSE) | Overall prediction error. |
| Pearson Correlation | Linear association between methods. |
| Intraclass Correlation Coefficient (ICC) | Agreement between measurements. |
| Bland–Altman Analysis | Agreement and systematic bias. |
Outputs
- Variable comparison tables.
- Agreement statistics.
- Validation summary.
- Technical recommendations.
Related Documentation
- TR-002 Processing Workflow
- TR-003 Variable Dictionary
- TR-005 Output Specification