Variations of One Factor ANOVA Designs - Agricultural Statistical Data Analysis Using Stata - page 94

var|maturity | 12.4521435 42 .296479608 3.37 0.0000
|
Residual | 7.38259695 84 .087888059
-------------+--------------------------------------------------
Total | 29.3538965 134 .219058929
The maturity groups are clearly different with an F value of
10.61. The maturity by replication interaction ( maturity#rep ,
0.158193325) is the mean square error used as the denominator to
calculate this value. This is accomplished in the command by using
the / character between maturity and rep#maturity .
The variety within maturity group sum of squares (12.4521435)
needs to be partitioned for each maturity group so that an accu-
rate F value can be calculated for each. The residual mean square
(0.087888059) is the correct term to use for the denominator in cal-
culating these F values, so at this point this value should be stored in
a macro. Enter the following command:
local x = e(rmse)^2
local y = e(df_r)
Remember the Root MSE (0.296459) is the square root of the
Residual mean square (0.087888059), which is the value we are inter-
ested in saving for future calculations. In addition, we are interested
in saving the Residual degrees of freedom e(df_r), which in this case
is 84. Now, enter the following command:
anova yield var rep if maturity == 1
This calculates an ANOVA based on the first maturity class and
results in the following output:
Number of obs = 45 R-squared = 0.7668
Root MSE = .262096 Adj R-squared = 0.6336
Source | Partial SS df MS F Prob > F
-------+----------------------------------------------------
Model | 6.3255123 16 .395344519 5.76 0.0000
|
var | 4.15479519 14 .296771085 4.32 0.0005
rep | 2.17071711 2 1.08535855 15.80 0.0000
|
Residual | 1.92344273 28 .068694383
---------+----------------------------------------------------
Total | 8.24895503 44 .187476251