SCIOVANTA

STATISTICS / INDEPENDENT GROUPS

ANOVA & p-values

Compare group means, inspect assumptions and export the complete statistical record.

  1. 1Data
  2. 2Settings
  3. 3Results
  4. 4Export

Data & settings

Synthetic example data

Columns: response,factor1. Add factor2 for two-way analysis. Supply independent observations, not group means. Up to 12 groups/cells and 5000 rows.

Analysis settings

Confidence intervals are fixed at 95%. No statistical method is switched automatically.

Results

Preparing the labelled synthetic example…

Method

F = MS between / MS within
p = Pr(F distribution ≥ observed F)

One-way ANOVA assumes independent observations, a common residual variance and an appropriate residual distribution. Welch ANOVA permits unequal group variances. Two-way analysis includes interaction and uses Type III partial sums of squares with sum-to-zero contrasts; all factor combinations require replication. A p-value is not the probability that the null hypothesis is true and does not measure practical importance.

NIST: one-way ANOVA · NIST: Levene/Brown–Forsythe · SciPy: classical and Welch definitions · statsmodels: Type III ANOVA

Additional methods and scope

This browser module computes classical/Welch/two-way ANOVA, variance tests, confidence intervals, effect sizes where defined, Holm-adjusted pairwise Welch tests and residual diagnostics. It does not compute Shapiro–Wilk, Tukey, Games–Howell, Dunnett or Dunn. Those methods remain in the original Python workbench, which requires a configured Python service. Repeated-measures and mixed-effects models are outside this module’s scope.

Export

Download full-precision results and inputs, or save a record in My work. Figure exports are beside each plot.