StuartMaxwellTest Class

public sealed class StuartMaxwellTest : HypothesisTest

Public NotInheritable Class StuartMaxwellTest
	Inherits HypothesisTest

public ref class StuartMaxwellTest sealed : public HypothesisTest

[<SealedAttribute>]
type StuartMaxwellTest = 
    class
        inherit HypothesisTest
    end

Inheritance

Object  →  HypothesisTest  →  StuartMaxwellTest

Use the StuartMaxwellTest to test for the significance of the marginal homogeneity of the columns in a K by K frequency table.

The null hypothesis of the Stuart-Maxwell test is that the marginal distributions of the rows and columns are the same. The test statistic follows a chi-square distribution with K-1 degrees of freedom.

This test is used in applications where you need to compare the marginal distributions of two related samples, such as in repeated measures or matched-pairs designs.

The data for the test must be provided as ICategoricalVector objects in the constructor. The CategoryIndex of each variable must have the same number of levels. The category indexes will often be the same, but this is not necessary. The two samples must also have the same number of observations.

If the categorical variables have only two categories, this test is equivalent to the McNemarTest.

This example demonstrates how to perform a Stuart-Maxwell test with two categorical vectors and then access its properties and methods.

using Numerics.NET.Statistics;
using Numerics.NET.Statistics.Tests;
using Numerics.NET.DataAnalysis;

// Create two categorical vectors with non-trivial inputs
ICategoricalVector sample1 = Vector.CreateCategorical(new[] { 1, 2, 1, 2, 3 });
ICategoricalVector sample2 = Vector.CreateCategorical(new[] { 2, 1, 2, 3, 1 });

// Perform the Stuart-Maxwell test
var test = new StuartMaxwellTest(sample1, sample2);

// Access the test statistic
double statistic = test.CalculateStatistic();
Console.WriteLine($"Test Statistic: {statistic}");

// Access the p-value
double pValue = test.GetPValue(HypothesisType.OneTailedUpper);
Console.WriteLine($"P-Value: {pValue}");

// Check if the null hypothesis is rejected at the 0.05 significance level
bool reject = test.Reject(0.05);
Console.WriteLine($"Reject Null Hypothesis: {reject}");

Reference