Statistics¶

P Value¶

A probability-scale test result used in many group comparisons. See also: Group comparisons.

An adjusted p value after multiple-testing correction. See also: Multiple testing.

False discovery rate, a common correction target for many simultaneous tests. See also: Multiple testing.

A scale of difference or association that is separate from the p value. See also: Effect sizes.

A standardized mean-difference effect size with small-sample correction. See also: Effect sizes.

A range estimate around a statistic or effect. See also: Effect sizes.

Whether a distribution is close enough to normal for parametric assumptions. See also: Normality outputs.

A test used when parametric assumptions are not suitable. See also: Statistics options.

A follow-up comparison after an omnibus or multi-group analysis. See also: Group comparisons.

Adjustment for testing many columns, groups, or contrasts. See also: Multiple testing.

A variable included to adjust or explain an outcome in a model. See also: Linear models.

A model-derived group mean after accounting for covariates or design terms. See also: Linear models.

A fitted model term describing direction and size of association. See also: Linear models.

The difference between an observed value and the model-fitted value. See also: Linear models.

A repeated train/test splitting method used to estimate model performance. See also: Classification.

Area under the receiver operating curve for classification. See also: Classification.

Accuracy averaged across classes so class imbalance has less influence. See also: Classification.

An F1 score averaged across classes. See also: Classification.

A test that compares the observed result to results after shuffled labels or assignments. See also: Classification.

The chance of detecting an effect under a planned design and assumed effect size. See also: Power.

A sinusoidal model for rhythmic data. See also: Rhythm statistics.

The phase or timing of a rhythm peak. See also: Rhythm statistics.