PRACTICAL CHECKLIST
A step-by-step approach.
Document your dataset
Record provenance, collection methods, units and permissions.
Inspect quality
Assess missing values, duplicates, outliers and unrealistic records.
Choose the analysis for the question
Select tests or models using assumptions, design and the inference you need.
Avoid leakage and overfitting
Separate evaluation data from model development where appropriate.
Report uncertainty
Include limitations, confidence intervals or other uncertainty measures and reproducible steps.
This guide is educational information, not a guarantee of academic, publication or application outcomes.
RELATED SERVICES