Scribd
Upload
107 views11 pages

ANOVA One Way

1. ANOVA models are used to compare the mean response values at different levels of factors in experimental studies. Each level of a factor is analyzed to see if the response is significantly different from other levels. 2. Analysis of variance partitions the total variance into between-groups variance and within-groups variance. If the between-groups variance is significantly higher than the expected within-groups variance, the null hypothesis that the means are equal is rejected. 3. An example is provided where four diets are compared using ANOVA. The calculated F value is higher than the critical F value so the null hypothesis is rejected, indicating at least one diet mean is significantly different from the others. Diet D

Uploaded by

Rahul Goyal
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
107 views11 pages

ANOVA One Way

1. ANOVA models are used to compare the mean response values at different levels of factors in experimental studies. Each level of a factor is analyzed to see if the response is significantly different from other levels. 2. Analysis of variance partitions the total variance into between-groups variance and within-groups variance. If the between-groups variance is significantly higher than the expected within-groups variance, the null hypothesis that the means are equal is rejected. 3. An example is provided where four diets are compared using ANOVA. The calculated F value is higher than the critical F value so the null hypothesis is rejected, indicating at least one diet mean is significantly different from the others. Diet D

Uploaded by

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

ANOVA

The analysis of experimental


studies involves the use of
analysis of variance (ANOVA)
models. ANOVA models are used
to compare the mean response
values at different levels of the
factor. Each level of the factor is
investigated to see if the
response is significantly different
from the response at other levels
of the factor. Analysis of variance
is a systematic approach towards
partitioning the total variance in
to between the groups and
within groups.
ANOVA
• F distribution is related to the Chi-square distribution
• X1 is a random variable and has Chi-square with v1 d.f
• X2 is a random variable and has Chi-square with v2 d.f
• New random variable = (X1/v1)/(X2/v2)= F calculated value
• F distribution often arises when we are working with ratio’s of variances
• Dependent variable – Ratio scale
• Normality
• Homoscedasticity
BASIC FROAME WORK OF ANOVA
ANOVA = MSB/MSW

MSB=SSB/Degrees of freedom between

MSW=SSW/Degrees of freedom within

 N Y  2
SSB = k Y

  Y  Y 
2
SSW=
Null hypothesis: All means are equal
H0 : Mean1 = Mean2 = Mean3 = Mean4
Alternate hypothesis: At-least one pair of means are
not equal.
H1 : Mean1 ≠ Mean2 ≠ Mean3 ≠ Mean4

If the variation within a group


is significantly different from
the variation between the
groups , then we would expect
to reject the null hypothesis
Example
An experiment was conducted to D1 D2 D3 D4
investigate the effects of four different
diets D1, D2, D3, and D4 on daily gains 30 32 34 36
(g) in weight of six chicks of 6 weeks old. 27 28 25 30
The following data are related to gain in 23 27 26 28
weights. Analyze the data and test
whether all the four diet treatments are 25 29 27 33
equally efficient and if not which diet is 23 26 25 32
the best 26 29 29 30
D1 D2 D3 D4
Between group variance: 30 32 34 36
27 28 25 30
1. Calculate grand mean for all
observations across all groups 23 27 26 28
2. Subtract each grand mean from each 25 29 27 33
sample mean
3. Square these difference 23 26 25 32
4. Multiply each squared score by 26 29 29 30
sample size.
5. Add them all up
Mean 25.67 28.50 27.67 31.50

 N Y  2 GM 28.33
6. Then SSB = k Y SSDM 7.11 0.03 0.44 10.03
SSB 105.66
Within group variance:
D1 D2 D3 D4
1. Calculate the mean for each 18.77778 12.25 40.11111 20.25
group 1.777778 0.25 7.111111 2.25
2. Subtract each sample mean from
7.111111 2.25 2.777778 12.25
every score in that group
3. Square the difference 0.444444 0.25 0.444444 2.25
4. Add up all the squared 7.111111 6.25 7.111111 0.25
differences 0.111111 0.25 1.777778 2.25

  Y  Y 
2
35.33 21.50 59.33 39.50
5. Then SSW =
  155.67   
Decision values

• MSB= SSB/d.f = 105.66/3=35.22


• MSW=SSW/d.f= 155.67/ 20=7.78
• F=MSB/MSW = 35.22/7.78 =4.53 F calculated value
• F table value= 3.10
• F- Calculated value is > than F table value, Reject the null hypothesis
Critical difference C.D

• If the difference between


any pair of means is
greater than the critical
difference value at
specified level of
significance, then the
means under comparison
differ significantly.
Decision among the diets
Diet D4 D2 D3 D1
M 31.50 28.50 27.67 25.67

Comparison between diets MD CD Conclusion


D4 & D2 3 3.35 No difference

D4 & D3 3.83 3.35 Difference is significant

D4 & D1 5.83 3.35 Difference is significant

D2 $ D3 0.83 3.35 No difference

D2 & D1 2.83 3.35 No difference

D3 & D1 2.00 3.35 No difference


Example 2
Feed1 Feed2 Feed3 Feed4
Thirty-two animals were 12.5 13.7 11.6 14.8
fed with four different 12 13.8 11.8 14.5
feeds. The following 12.3 13.9 11.9 14.9
figures give the gain in 13.4 13.8 12.2 15
body weight after 2 13.5 14 11.8 14.7
months. Analyze the data
13.6 12.9 12.1 14.5
and draw your conclusion
11.9 13.5 12.4 14.9
12.7  11.9 14.8
12.8  11.9 
12.4     
13.2     

You might also like