stats

F-Test for Variances

Compare two sample variances — F statistic, p-value, F critical value, ratio CI, and F-distribution curve.

⚠ The F-test for variances is highly sensitive to non-normality. Consider Levene's Test for non-normal data.

Inputs

Input mode

Group 1 (comma-separated)Group 2 (comma-separated)

Hypothesis tail

Results

Group 1 variance s₁²—

Group 2 variance s₂²—

df₁ / df₂—

F statistic (G1/G2)—

p-value (two-tailed)—

Variance ratio s₁²/s₂²—

Interpretation—

Conclusion α = 0.05—

Conclusion α = 0.01—

F = s²_larger / s²_smaller (larger always in numerator)
p via P(F>f) = 1 − I_x(df₁/2, df₂/2) x = df₁·f/(df₁·f + df₂)

I_x via Lentz continued-fraction method (OneWayAnova.tsx pattern).
Reliable for df ≤ ~100.