Thanks to visit codestin.com
Credit goes to www.scribd.com

Scribd
Upload
104 views26 pages

Nonparametric Statistical Tests Guide

This document provides an overview of nonparametric and distribution-free statistical methods, including the sign test, Wilcoxon signed-rank test, Wilcoxon rank-sum test, and Kruskal-Wallis test. These tests make fewer assumptions about the underlying population distribution compared to parametric tests. Examples are provided to demonstrate how to perform and interpret the results of each test. The document also notes that nonparametric tests can be more powerful than parametric tests when the normality assumption is violated, and the efficiency of nonparametric tests relative to parametric tests tends to be higher for distributions with heavier tails than the normal.

Uploaded by

ryanroncales
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
104 views26 pages

Nonparametric Statistical Tests Guide

This document provides an overview of nonparametric and distribution-free statistical methods, including the sign test, Wilcoxon signed-rank test, Wilcoxon rank-sum test, and Kruskal-Wallis test. These tests make fewer assumptions about the underlying population distribution compared to parametric tests. Examples are provided to demonstrate how to perform and interpret the results of each test. The document also notes that nonparametric tests can be more powerful than parametric tests when the normality assumption is violated, and the efficiency of nonparametric tests relative to parametric tests tends to be higher for distributions with heavier tails than the normal.

Uploaded by

ryanroncales
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 26

Bootstrap Method

15-1 Introduction

•  Most of the hypothesis-testing and confidence


interval procedures discussed in previous chapters
are based on the assumption that we are working
with random samples from normal populations.
•  These procedures are often called parametric methods
•  In this chapter, nonparametric and distribution free
methods will be discussed.
•  We usually make no assumptions about the distribution
of the underlying population.
15-2 Sign Test
15-2.1 Description of the Test
•  The sign test is used to test hypotheses about the
median of a continuous distribution.

• Let R+ represent the number of differences


~
X i − µ0
that are positive.
•  What is the sampling distribution of R+ under H0?
15-2 Sign Test
15-2.1 Description of the Test

If the following hypotheses are being tested:

The appropriate P-value is


15-2 Sign Test

15-2.1 Description of the Test


If the following hypotheses are being tested:

The appropriate P-value is


15-2 Sign Test
15-2.1 Description of the Test
If the following hypotheses are being tested:

If r+ < n/2, then the appropriate P-value is

If r+ > n/2, then the appropriate P-value is


15-2 Sign Test
Example 15-1
Example 15-1
15-2 Sign Test
Example 15-1
15-2 Sign Test
15-2.2 Sign Test for Paired Samples

See Example 15-3.


15-2 Sign Test
15-2.3 Type II Error for the Sign Test
•  Depends on both the true population distribution and alternative value!

Figure 15-1
Calculation of β
for the sign test.
(a) Normal
distributions. (b)
Exponential
distributions
15-3 Wilcoxon Signed-Rank Test

•  The Wilcoxon signed-rank test applies to the case


of symmetric continuous distributions.
•  Under this assumption, the mean equals the median.
•  The null hypothesis is H0: µ = µ0
15-3 Wilcoxon Signed-Rank Test
•  15-3.1 Description of the Test
•  Assume that X1, X2, …, Xn is a random sample from a continuous
and symmetric distribution with mean (and median) µ.
Procedure:
•  Compute the differences Xi − µ0, i = 1, 2, …, n.
•  Rank the absolute differences |Xi − µ0|, i = 1, 2, …, n in ascending
order.
•  Give the ranks the signs of their corresponding differences.
•  Let W+ be the sum of the positive ranks and W− be the absolute
value of the sum of the negative ranks.
•  Let W = min(W+, W−).
15-3 Wilcoxon Signed-Rank Test
Decision rules:
15-3 Wilcoxon Signed-Rank Test
Example 15-4
Example 15-4
15-3 Wilcoxon Signed-Rank Test
Example 15-4
15-3 Wilcoxon Signed-Rank Test

15-3.2 Large-Sample Approximation

Z0 is approximately standard normal when n is large.


15-4 Wilcoxon Rank-Sum Test

15-4.1 Description of the Test


15-4 Wilcoxon Rank-Sum Test
15-4.1 Description of the Test
15-4 Wilcoxon Rank-Sum Test
Example 15-6
15-4 Wilcoxon Rank-Sum Test
Example 15-6
Example 15-6
15-4 Wilcoxon Rank-Sum Test
Example 15-6
15-5 Nonparametric Methods in the
Analysis of Variance

The single-factor analysis of variance model for


comparing a population means is

The hypothesis of interest is

The Kruskal-Wallis test (w/o assumption of normality)


•  Basic idea: Use ranks instead of actual numbers
Parametric vs. Nonparametric Tests
•  When the normality assumption is correct, t-test or F-test is
more powerful.
–  Wilcoxon signed-rank or rank-sum test is approximately 95% as
efficient as the t-test in large samples.
•  On the other hand, regardless of the form of the
distributions, nonparametric tests may be more powerful.
–  Wilcoxon signed-rank or rank-sum test will always be at least 86%
as efficient.
•  The efficiency of the Wilcoxon test relative to the t-test is
usually high if the underlying distribution has heavier tails
than the normal
–  because the behavior of the t-test is very dependent on the sample
mean, which is quite unstable in heavy-tailed distributions.

You might also like