Professional Documents
Culture Documents
ANOVA
ANOVA
1
11/24/2017
This material dealing in two-sample inference In the k > 2 sample problem, it will be assumed that
represents a special case of what we call the one-factor
problem. there are k samples from k populations. One very
common procedure used to deal with testing
For example, the survival time was measured for two population means is called the analysis of
samples of mice, where one sample received a new variance, or ANOVA.
serum for leukemia treatment and the other sample
received no treatment. In this case, we say that there is
one factor, namely treatment, and the factor is at two Analysis of variance (ANOVA) is a method for
levels. If several competing treatments were being used
in the sampling process, more samples of mice would be splitting the total variation of the data into
necessary. In this case, the problem would involve one meaningful components that measure different
factor with more than two levels and thus more than two sources of variation.
samples.
2
11/24/2017
3
11/24/2017
Two Sources of Variability in the Data Two Sources of Variability in the Data
4
11/24/2017
5
11/24/2017
3. The variables from each group come from 1. Determine the null (H0) and alternative hypothesis
distributions with approximately the same standard (H1)
deviation. H0: μ1 = μ2 = . . . = μk
6
11/24/2017
7
11/24/2017
8
11/24/2017
A set of observations may be classified according Also, to determine if part of the variation is due to
to two criteria at once by means of a rectangular differences among the columns, we consider the test
array in which the columns represent one criterion of
classification and the rows represent a second H0”: μ1 = μ2 = . . . = μ.c = μ
criterion of classification. H1’: the μj are not all equal
To determine if part of the variation in our
observations is due to differences among the rows, The sum-of-squares identity may be represented
we consider the test symbolically by the equation
H0’: μ1 = μ2 = . . . = μr. = μ
H1’: the μi are not all equal SST SSR SSC SSE
9
11/24/2017
where:
r c
T..2
SST xij2
i 1 j 1 rc
r
T 2
i.
T..2
SSR i 1
c rc
c
T
j 1
2
.j
T..2
SSC
r rc
SSE SST SSR SSC
10
11/24/2017
Example: Example
The following data represent the yields of the
three varieties of wheat, using four different kinds of Test the hypothesis H0’ that there is no
fertilizer. difference in the average yield of wheat when
different kinds of fertilizer are used. Also, test the
hypothesis H0” that there is no difference in the
average yield of the three varieties of wheat. Use
0.05 level of significance.
11
11/24/2017
Exercise:
12
11/24/2017
13