Two-sample variance F-test

When to consider it

• Compare variances of two independent normal populations.

Assumptions and scope

• Normal populations are required for exact inference. These tests are sensitive to non-normality; consider median-centered Levene for comparing groups.

Required inputs

• Sample 1 — Numeric observations; preserve the row and column structure described above.
• Sample 2 — Numeric observations; preserve the row and column structure described above.

Review missing values and data coding before running. Sample inputs are synthetic teaching examples.

Implementation

Two-sided F-test of variance ratio 1 with a 95% ratio interval. Both populations must be normal and samples independent. Levene is a less normality-sensitive alternative. Synthetic example: independent cohorts of 40 and 35 simulated measurements with different variability.

• scipy 1.15.3: scipy.stats.f.cdf, scipy.stats.f.sf, scipy.stats.f.ppf

These are pinned library versions; saved results record the versions actually used at runtime.

Library references

• SciPy statistical methods
• statsmodels models and statistical tests
• scikit-learn clustering
• lifelines survival comparisons