MultipleTesting.AdjustHochberg Method

public static Vector<double> AdjustHochberg(
	Vector<double> pValues
)

Public Shared Function AdjustHochberg ( 
	pValues As Vector(Of Double)
) As Vector(Of Double)

public:
static Vector<double>^ AdjustHochberg(
	Vector<double>^ pValues
)

static member AdjustHochberg : 
        pValues : Vector<float> -> Vector<float> 

Parameters

pValues  Vector<Double>

A vector of raw p-values.

Return Value

This method applies the Hochberg correction to control the family-wise error rate.

The procedure works by ordering p-values from largest to smallest, then applying a step-up procedure that adjusts p-values based on their rank order.

The Hochberg procedure is generally more powerful than the Holm procedure while still controlling the family-wise error rate at the desired level. It is valid for independent or positively correlated tests.

This procedure requires that p-values be ordered in descending order for proper application. The implementation handles this sorting internally. Additionally, if the tests have negative dependencies, this method may not control the error rate appropriately.

Reference