Scribd
Upload
53 views

14 F DISTRIBUTION - Pps

The F-distribution is used to determine if the variances of two populations are significantly different. It compares the variance of two samples by taking the ratio of their variance estimates. This ratio, F, follows an F-distribution with degrees of freedom based on the sample sizes. An F-value close to 1 means the variances are likely the same, while a very large or small F means the variances are likely different. By comparing the calculated F to critical F-values, one can test if variances are statistically different at a given significance level, such as 5%.

Uploaded by

Arya Pradana
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPS, PDF, TXT or read online on Scribd
53 views

14 F DISTRIBUTION - Pps

The F-distribution is used to determine if the variances of two populations are significantly different. It compares the variance of two samples by taking the ratio of their variance estimates. This ratio, F, follows an F-distribution with degrees of freedom based on the sample sizes. An F-value close to 1 means the variances are likely the same, while a very large or small F means the variances are likely different. By comparing the calculated F to critical F-values, one can test if variances are statistically different at a given significance level, such as 5%.

Uploaded by

Arya Pradana
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPS, PDF, TXT or read online on Scribd
You are on page 1/ 11

The F-Distribution

1 F-Distribution
The F-Distribution

In studying the distribution of the difference between means that the
F-distribution is used to determine the validity of the assumption of
identical standard deviations of two populations

Used in the analysis of variance (Anova), will discuss next


It has two numbers of degree of freedom (the denominator
and the numerator of the degree of freedoms)

y
If F(v1, v2) satisfies the F-
distribution, then:

1
F  v2 , v1  
'

F  v1 , v2 
satisfies the F-distribution
1 F also
2 F-Distribution
Testing the homogeneity of Two
Variances

The variance of two samples are said to be homogeneous if the


population from which the two samples are taken have identical
variances (or identical standard deviation)
First population, with means:
X¯1, X¯2, …. X¯N
Second population, with means:
Y¯1, Y¯2, …… YN
Calculate the best estimate of the standard deviation of the
population

The best estimate obtained from the samples of the first


population are s x,1 s x,2 ……..s x,N and those from second
population are s y,1 s y,2 ….. S y,N

3 F-Distribution
That is

( X  X 1 ) 2 (Y  Y 1 ) 2
s x ,1  s y ,1 
n1  1 n2  1

Calculate for each pair the ratio of the squares of


the best estimates of the standard deviation of two
samples, thereby N such ratios are obtained which
we shall denote by F1, F2, …. FN as shown in figure
in the next slide

4 F-Distribution
Ratio of squares of
best estimates of
standard deviations
Mean = X¯1 Mean = Y¯1
No.Variates = n1 No.Variates = n2 s x2,1
Best est.of.pop. Best est.of.pop. F1 
st.dev=sx,1 st.dev= sy,2 s y2,1

First Mean = y¯2


population
Mean = X¯2 No.Variates = n2 s x2, 2
No.Variates = n1
Best est.of.pop. F2 
X1, X2 … XN1 Second
Best est.of.pop.
st.dev=sx,2
st.dev= sy,2 s y2, 2
population
Y1, Y2, ,.. YN2

Mean = X¯N Mean = y¯N


No.Variates = n1 No.Variates = n2 s x2, N
Best est.of.pop. Best est.of.pop. FN 
st.dev=sx,N st.dev= sy,N s 2y , N
The F values thus obtained satisfy an F-Distribution with v1 = n1 – 1
and v2 = n2 -1
Theorem:
If X and Y are normally distributed with identical standard
deviations, and if all possible pairs of samples are drawn from
these populations (as shown in figure of previous slide) then the
ratio of the squares of the best estimates of the standard
deviations, that is

s x2
F  2
sy
Where :

( X  X 1 ) 2 (Y  Y 1 ) 2
sx  sy 
n1  1 n2  1

Satisfy an F-distribution with v1 = n1 – 1 and v2= n2 - 1


6 F-Distribution
From this theorem, it follows immediately that, for the
arrangement described as in previous slide, also the
reciprocal of F, that is

s y2
F 
s x2
Satisfy the F-distribution with v1 = n1 -1, and v2 = n2 -1

To test the homogeneity of variances, we set up the hypothesis that the standard
deviation of the two populations under consideration are identical. Thus the value of F
as obtained is expected to be near 1. If a very large or very small value of F is obtained, it
would be an indication that the hypothesis of equal standard deviations should be
rejected.
7 F-Distribution
A two-tailed test is always applied in this case, since the problem is to
find whether sx and sy differ significantly. We must be prepared to reject
the hypothesis of identical population standard deviation if F is either
significantly larger or significantly smaller than 1.
Here the hypothesis σx = σy is considered as having the alternative σx
≠ σy

In many other problems, one variance is assumed given, and the


question is raised whether another variance exceeds the given
one by a significant amount. This situation call for one-tailed test,
as the alternative to the hypothesis σx = σy then is σx > σy where σy
is given or fixed variance.

The statistical table of F- distribution gives the critical F-value for a given level
of significance, and we compare it with the calculated F-value for getting a
conclusion in homogeneity test
8 F-Distribution
From the data (see student-t example)
X : 138, 127, 134, 125
Y: 134, 137, 135, 140, 130, 134
determine whether the variances of the two samples can be
considered homogeneous on the basis of 5% level of
significance?
 ( X  X ) 2
110
s x2    36.67
n1  1 3
 (Y  Y ) 2
56
s y2    11 .20
n2  1 5
36.67
and F   3.27
11 .20

From statistical table we find F0.975 (3,5)= 7.76  we accept the


hypothesis
9 F-Distribution
THANK YOU

10 F-Distribution
11 F-Distribution

You might also like