In: Statistics and Probability
Consider the following partially completed two-way ANOVA table. Suppose there are 2 levels of Factor A and 3 levels of Factor B. The number of replications per cell is 3. Use the 0.01 significance level. (Hint: estimate the values from the F table.)
SS df MS F
Factor A 100
Factor B 30
Interaction 250
Error 200
Total 580
(a)
The ANOVA Table is completed asfollows:
Source of Variation | Sum of Squares | Degrees of freedom | Mean Square | F |
Factor A | 100 | a - 1= 2 - 1 = 1 | 100/1 = 100 | 100/16.6667=6.00 |
Factor B | 30 | b - 1 = 3 - 1 = 2 | 30/2 = 15 | 15/16.6667=0.900 |
Interaction | 250 | (a - 1) X (b - 1) = 1 X 2 = 2 | 250/2 =125 | 125/16.6667=7.500 |
Error | 200 | ab(r-1) = 2 X 3 X 2 = 12 | 200/12 = 16.6667 | |
Total | 580 | 17 |
(b)
To find the critical values to test for equal means:
(i)
= 0.01
For Factor A:
Degress of freedom for numerator = 1
Degress of freedom for denominator = 12
From Table:
Critical value of F = 9.33
(ii)
= 0.01
For Factor B:
Degress of freedom for numerator = 2
Degress of freedom for denominator = 12
From Table:
Critical value of F = 6.927
(iii)
= 0.01
For Interaction:
Degress of freedom for numerator = 2
Degress of freedom for denominator = 12
From Table:
Critical value of F = 6.927
(c)
(i)
For Factor A:
Since calculated value of F = 6.00 is less than critical value of F = 9.33, there is no significant difference in Factor A means,
(ii)
For Factor B:
Since calculated value of F = 0.900 is less than critical value of F = 6.927, there is no significant difference in Factor B means,
(d)
For Interaction:
Since calculated value of F = 7.500 is greater than critical value of F = 6.927, there is a significant difference in Interaction means,