When to use this
Use this to understand the sample before testing research claims.
Why it matters
Descriptive evidence reveals coding problems and provides context for every later model.
Step-by-step instructions
1. Open Analysis and choose Find an analysis.
2. Search for descriptive statistics, or browse Descriptive & Exploratory.
3. Confirm the active project, dataset, and immutable version.
4. Review the automatically generated preview and its missing-data handling.
5. Adjust contextual settings; temporary results refresh without creating a research record.
6. Save an immutable analysis only after the evidence is ready for history and reporting.
7. Open Results & Reports to reuse the saved analysis in a Results Notebook.
Expected outputs
Version-bound tables, charts, diagnostics, and reproducibility metadata.
How to interpret the results
Describe the observed sample and distribution. Avoid turning descriptive differences into causal or population claims.
Common mistakes
Do not ignore missing denominators. Do not report excessive decimal precision or treat a mean as sufficient for every variable.
Was this article helpful?
Your response helps the methodology team improve this documentation.