OriginLab Corporation - Data Analysis and Graphing Software - 2D graphs, 3D graphs, Contour Plots, Statistical Charts, Data Exploration, Statistics, Curve Fitting, Signal Processing, and Peak Analysis                     
 
Skip Navigation Links
All BooksExpand All Books
Origin HelpExpand Origin Help
StatisticsExpand Statistics
Nonparametric Tests (Pro Only)Expand Nonparametric Tests (Pro Only)
Friedman ANOVAExpand Friedman ANOVA

17.5.8 Friedman ANOVA

Introduction

The Friedman ANOVA is a nonparametric test that compares three or more paired groups by determining the score difference between k treatments of n blocks. It performs a nonparametric analysis of a randomized block experiment and thus provides an alternative to the one-way repeated measure ANOVA.

This test requires data to follow a balanced design, i.e. exactly one observation per treatment-block combination.

The hypothesis under test,H_0\,\! , often called the null hypothesis, is that the k samples come from the same population, and this is to be tested against an alternative hypothesis H_1\,\!that they come from different populations.

Handling Missing Values

Friedman ANOVA does not support data with missing values. Please trim your missing values from data and make sure the data follow balanced design before running Friedman ANOVA

Performing Friedman ANOVA

To perform a Friedman ANOVA:

  1. Select Statistics: Nonparametric Tests: Friedman ANOVA. This opens the friedman dialog box.
  2. Specify the Input Data.
  3. Upon clicking OK, an analysis report sheet is generated that lists degrees of freedom, Chi-square statistics, the associated p-value, and the test conclusion.


Topics covered in this section:

 

© OriginLab Corporation. All rights reserved.