- The setting SignificanceLevel->α with 0<α<1 indicates a significance level of α.
- The level used for verifying assumptions such as normality is taken to be where is the number of datasets being tested.
Examplesopen allclose all
Basic Examples (2)
Set test sizes for diagnostics used in LocationTest:
The -value is equivalent but a warning is issued:
Set the cutoff used in "TestConclusion" and "ShortTestConclusion" properties:
By default, 0.05 is used:
Properties & Relations (2)
The significance level can affect the degrees of freedom for some tests:
An assumption of equal variances is rejected at the 25% level and a Satterthwaite approximation is used for the degrees of freedom:
The significance level is divided by the number of comparisons when assumptions are checked:
Using with 50 tests for normality results in each one being tested at the level: