Two and More Factors ANOVA - Agricultural Statistical Data Analysis Using Stata - page 121

Predictive Margins of Fertility#Sowingdate with 95% Cls

100

80

60

40

20

5 Oct. 2001
29 Oct. 2001

15 Oct. 2001

0

0 50

Fertility: 0-200 lbs/acre nitrogen

100 150 120

Predictive Margins of Variety#Sowingdate with 95% Cls

80

60

40

20

5 Oct. 2001
29 Oct. 2001

15 Oct. 2001

0
Nirvana Pegasus

Variety: 1-Nirvana, 2-Pegasus, 3-PS 7092, 4-Sweet Vidalia

PS 7092 Sweet Vidalia

Figure 6.7 The first graph using the marginsplot command after margins
fertility#sowingdate and the second graph using the marginsplot command
after the margins variety#sowingdate.

be seen in this example where the error degrees of freedom increases
from the main plot, to the subplot, and finally the sub-subplot.
Because the error sum of squares are divided by the error degrees of
freedom and this is used as the denominator in an F-test, it is easy to
see how the precision and ability to detect differences would increase.
The arrangement of this design assigns first the main-plot effects ran-
domly to the largest unit within the experiment. The subplot factors
are then assigned randomly within the main plots and finally the sub-
subplot factors are randomly assigned to within the subplots.
Figure 6.8 shows the layout of a split-split plot design with three
replications arranged with five main-plot nitrogen fertility treatments,