Variance test

Returns the P-value of a test that tests the variance of the data against a particular value.

Decision	Reality: H0 is correct	Reality: H1 is correct
Accept H0	No problem	Type 1 error
Reject H0	Type 2 error	No problem

Definition

P-value is basically a probability of making a Type 2 error.

VarTest returns the P-value for the hypothesis test in the following form:

• H0:

  

• H1:

  

  

  

  

Syntax

VarTest <Hypothesis Type> (Argument, Hypothesis Variance)

Where:

Hypothesis Type is the parameter to set whether you are testing the probability of the value being equal (0), less than (-1), greater than (1), or not equal to (2) the sigma or variance.

Argument is a fact or metric representing the values you are testing.

HypothesisVariance is the variance you are testing.

Example

VarTest(M1, 1000) = 1000

For a set of values (M1), what is the probability that the variance of the value will be equal to 1000? Note that M1 needs to have a defined dimensionality.

Usage Notes

For each variance test function, two tests are performed. The first tests the probability of the value being equal (0), and the second tests the probability of the value being less than (-1), greater than (1), or not equal to (2) the sigma or variance. A parameter must be specified for the second test.