Scribd
Upload
16 views19 pages

Chi 2 B

The document discusses the Chi-square goodness of fit test, providing examples to test hypotheses about the distribution of lift usage and random digits. It includes calculations for expected values, observed frequencies, and Chi-square statistics, leading to conclusions on whether null hypotheses can be rejected. Additionally, it covers a Chi-square test of independence related to employee attitudes and a survey of engineering students' opinions on a new building.

Uploaded by

jainarayancareer15
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
16 views19 pages

Chi 2 B

The document discusses the Chi-square goodness of fit test, providing examples to test hypotheses about the distribution of lift usage and random digits. It includes calculations for expected values, observed frequencies, and Chi-square statistics, leading to conclusions on whether null hypotheses can be rejected. Additionally, it covers a Chi-square test of independence related to employee attitudes and a survey of engineering students' opinions on a new building.

Uploaded by

jainarayancareer15
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/ 19

Chi-square goodness of fit test

Example 1
An office block has 4 lifts. The following table gives the number of
passengers per lift for a certain time interval.
Lift Number 1 2 3 4 Total
Number of Passengers 25 40 15 36 116
Test the hypothesis that all four lifts are equally busy over a long
time. (Use α = 0.05)
1. H0 = The lifts are equally busy. (Null Hypothesis.)
2. H1 = The lifts are not equally busy. (Alternative hypothesis)

Michael Carr Statistics


Chi-square goodness of fit test

Solution
Expected value = 116
4 = 29 (i.e. assuming H0 true, lifts equally
busy)
Lift Number 1 2 3 4 Total
Observed frequency 25 40 15 36 116
Expected frequency 29 29 29 29 116
First we must calculate the χ2 statistic
Test for Goodness of fit

(oi − ei )2
Σ( ) ≈ χ2n−1 (1)
ei

Michael Carr Statistics


Chi-square goodness of fit test

Then we must look up a value in the χ2 Distribution Tables. χ2


DISTRIBUTION TABLES
(Page 37 Department of Education Tables)
3 degrees of freedom-go to row 3
α = 0.05 so we read from the column =0.05
This gives us a value of 7.81

Michael Carr Statistics


Chi-square goodness of fit test

Rule We reject Ho if the chi 2 statistic is greater than the table


value
χ2 statistic (2)
(25 − 29)2 (40 − 29)2 (15 − 29)2 (36 − 29)2
= + + + (3)
29 29 29 29

0.55 + 4.17 + 6.76 + 1.69 = 13.17 (4)


Statistic = 13.17 > 7.81
So we Reject the Null hypothesis
Conclusion The lifts are not equally busy Since Statistic (13.17)
> Tables (7.81)

Michael Carr Statistics


Chi-square goodness of fit test

Exercise 1 Test the hypothesis that the following set of 200


numbers are random digits, that is, each number is equally likely to
be 0, 1, 2, 3, .......9.
(Use α = 0.05)
Digit 0 1 2 3 4 5 6 7 8 9
Frequency 22 16 15 18 16 25 23 17 24 24
Test the hypothesis that all numbers are equally likely over a long
time. (Use α = 0.05)

Michael Carr Statistics


Chi-square goodness of fit test

Solution
1. H0 = All the numbers are equally likely. (Null Hypothesis.)
2. H1 = All the numbers are not equallly likely. (Alternative
hypothesis)

Michael Carr Statistics


Expected value = 200
10 = 20 (i.e. assuming H0 true, lifts equally
busy)
Digit 0 1 2 3 4 5 6 7 8 9
Frequency 22 16 15 18 16 25 23 17 24 24
Expected Value 20 20 20 20 20 20 20 20 20 20

Michael Carr Statistics


Chi-square goodness of fit test

First we must calulate the χ2 statistic


Test for Goodness of fit
(oi − ei )2
Σ( ≈ χ2n−1 (5)
ei
Notation
1. oi = observed value for case i where i = 1, 2, 3, 4 etc
2. ei = expected value for case i where i = 1, 2, 3, 4 etc
3. Degrees of freedom = No. of columns - no. of constraints =
10 - 1 = 9 The constraint is that the total of expected values
must = 200

Michael Carr Statistics


Chi-square goodness of fit test

Then we must look up a value in the χ2 Distribution Tables.


χ2 DISTRIBUTION TABLES (Page 37 Department of Education
Tables)
3 degrees of freedom-go to row 9
α = 0.05 so we read from the column =0.05
This gives us a value of 16.92

Michael Carr Statistics


Chi-square goodness of fit test

We reject Ho if the chi 2 statistic is greater than the table value

(22 − 20)2 (16 − 20)2 (15 − 20)2 (18 − 20)2 140


χ2 statistic = + + + =
20 20 20 20 20
(6)

Michael Carr Statistics


Chi-square goodness of fit test

χ2 Statistic = 7 <16.92
So we accept the Null hypothesis Conclusion
Each Number is equally likely Since Statistic (7) < Tables (16.92)

Michael Carr Statistics


Exercise 2

The number of machine failures per day in a certain factory is


assumed to have a poisson distribution with parameter =0.5.
A count of the number of failures reveals the following.
No. of Failures 0 1 2 3
No. of days 65 25 8 2
Expected Value
Is the data consistent with the assumption at the 5 percent level of
significance ?
Note: Cells must be merged when an expected value is less than 5

Michael Carr Statistics


Chi Square test of Independence

Example 2
A random sample of 260 employees of a large multi-national
corporation participated in a survey and the employees were
classified by two characteristics, as shown in the table below.
Attitude to management Clerical Technician Sales Total
Satisfied 40 50 40 130
Dissatisfied 60 40 30 130
Total 100 90 70 260
Test the hypothesis that the two classifications are independent
(take α = 0.05). Give the degrees of freedom of the test statistic
and state your conclusion clearly.

Michael Carr Statistics


Solution

Ho = Attitude Towards Management is independent of Employee


Type (Null Hypothesis)
H1 = Attitude Towards Management is NOT independent of
Employee Type (Alternative Hyp.)

Michael Carr Statistics


Solution

Expected value of each cell row total×column


grandtotal
total

Example
Expected value of clerical and satisfied 130×100
260 = 50

Michael Carr Statistics


Solution

Attitude to management Clerical Technician Sales Total


Satisfied 40(50) 50 (45) 40(35) 130
Dissatisfied 60 (50) 40(45) 30(35) 130
Total 100 90 70 260

(oi − ei )2
Σ( ) ≈ χ2(c−1)(r −1) (7)
ei
Degress of freedom
=(no. of cols-1)(no. of rows-1)=(3-1)(2-1)=2

Michael Carr Statistics


Solution

χ2 DistributionTables 2 degress of freedom, α = 0.05,


Rule:
REJECT H0 if χ2statistic > 5.991

(40 − 50)2 (50 − 45)2 (40 − 35)2 (60 − 50)2


χ2statistic = + + + (8)
50 45 35 50
(40 − 45)2 (30 − 35)2
+ + = 6.54 (9)
45 35
Conclusion Since χ2statistic (6.54) > χ2tables (5.991)
Level of satisfaction is not Independent of Employee type.

Michael Carr Statistics


Exercise 1

A poll of 200 engineering students was taken to determine their


opinions concerning a new building. Each student was then
classified according to nationality and opinion on the new building.
The survey results are listed in the table below. Test the null
hypothesis that these two classifications are independent of one
another (that is, students do not hold opinions along national
lines) against the alternative hypothesis that students opinions on
the new building are related to their nationality. (use α = 0.10)
Nationality Approve Don’t No Opinion Total
French 40 20 10 70
Irish 60 50 20 130
Total 100 70 30 200

Michael Carr Statistics


Michael Carr Statistics

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy