0% found this document useful (0 votes)
41 views

Chapter 9 - Lecture 3 Tests About A Population Proportion: Andreas Artemiou

This document outlines a lecture on testing population proportions. It introduces testing one sample proportion for small and large samples. For large samples, it discusses the test statistic, rejection regions, and conditions for a valid test. It also covers calculating type II error probabilities and sample size. Examples are provided to illustrate the concepts and exercises are listed for practice.

Uploaded by

Ndomadu
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
41 views

Chapter 9 - Lecture 3 Tests About A Population Proportion: Andreas Artemiou

This document outlines a lecture on testing population proportions. It introduces testing one sample proportion for small and large samples. For large samples, it discusses the test statistic, rejection regions, and conditions for a valid test. It also covers calculating type II error probabilities and sample size. Examples are provided to illustrate the concepts and exercises are listed for practice.

Uploaded by

Ndomadu
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 10

Outline

Introduction
Testing one sample proportion for small sample
Testing one sample proportion for large sample
Exercises

Chapter 9 - Lecture 3
Tests about a population proportion

Andreas Artemiou

March 3rd, 2010

Andreas Artemiou Chapter 9 - Lecture 3 Tests about a population proportion


Outline
Introduction
Testing one sample proportion for small sample
Testing one sample proportion for large sample
Exercises

Introduction

Testing one sample proportion for small sample


Details through example

Testing one sample proportion for large sample


Details
Type II error and sample size

Exercises

Andreas Artemiou Chapter 9 - Lecture 3 Tests about a population proportion


Outline
Introduction
Testing one sample proportion for small sample
Testing one sample proportion for large sample
Exercises

Tests about population proportion

I There are cases where instead of having continuous datasets


we have discrete ones. In that case we are testing the
proportion of times something happens. (For example, the
proportion of times we have heads when we toss a coin)
I There are two type of tests in this case:
I Testing one sample proportion for small sample
I Testing one sample proportion for large sample

Andreas Artemiou Chapter 9 - Lecture 3 Tests about a population proportion


Outline
Introduction
Testing one sample proportion for small sample Details through example
Testing one sample proportion for large sample
Exercises

Example 9.13 page 446

Andreas Artemiou Chapter 9 - Lecture 3 Tests about a population proportion


Outline
Introduction
Details
Testing one sample proportion for small sample
Type II error and sample size
Testing one sample proportion for large sample
Exercises

Large sample test for one sample proportion

I Null Hypothesis: H0 : p = p0
p p0
I Test statistic: z = p
p0 (1 p0 )/n
I Rejection Regions:
I z z if HA : p > p0
I z z if HA : p < p0
I z z/2 and z z/2 if HA : p 6= p0
I There are two conditions that we need to check, so that the
use of this test is valid. We need:
I np0 10
I n(1 p0 ) 10

Andreas Artemiou Chapter 9 - Lecture 3 Tests about a population proportion


Outline
Introduction
Details
Testing one sample proportion for small sample
Type II error and sample size
Testing one sample proportion for large sample
Exercises

Example 9.11 page 443

Andreas Artemiou Chapter 9 - Lecture 3 Tests about a population proportion


Outline
Introduction
Details
Testing one sample proportion for small sample
Type II error and sample size
Testing one sample proportion for large sample
Exercises

Calculating Type II error probabilities

!
0+z
p
p0 p p 0 (1 p 0 )/n
I (p 0 ) = P z < p if HA : p > p0
p 0 (1 p 0 )/n
!
0z
p
p0 p p 0 (1 p 0 )/n
I (p 0 ) = 1 P z < p if
p 0 (1 p 0 )/n
HA : p < p0
!
0+z
p
p0 p /2 p 0 (1 p 0 )/n
I (p 0 ) = P z < p
p 0 (1 p 0 )/n
!
p0 p 0 z/2 p0 (1 p0 )/n
p
P z< p if HA : p 6= p0
p 0 (1 p 0 )/n

Andreas Artemiou Chapter 9 - Lecture 3 Tests about a population proportion


Outline
Introduction
Details
Testing one sample proportion for small sample
Type II error and sample size
Testing one sample proportion for large sample
Exercises

Calculating sample size

I If we have a test with size , type II error at p 0 equal to


(p 0 ) = is
" p p #2
z p0 (1 p0 ) + z p 0 (1 p 0 )
I n= if HA : p > p0 or
p 0 p0
H A : p < p0
" p p #2
z/2 p0 (1 p0 ) + z p 0 (1 p 0 )
I n= if HA : 6= 0
p 0 p0

Andreas Artemiou Chapter 9 - Lecture 3 Tests about a population proportion


Outline
Introduction
Details
Testing one sample proportion for small sample
Type II error and sample size
Testing one sample proportion for large sample
Exercises

Example 9.12 page 445

Andreas Artemiou Chapter 9 - Lecture 3 Tests about a population proportion


Outline
Introduction
Testing one sample proportion for small sample
Testing one sample proportion for large sample
Exercises

Exercises

I Section 9.3 page 446


I Exercises 36, 37, 38, 39, 40, 41, 42, 43, 44

Andreas Artemiou Chapter 9 - Lecture 3 Tests about a population proportion

You might also like