Variations of One Factor ANOVA Designs - Agricultural Statistical Data Analysis Using Stata - page 76

Many commands, including anova, have the general form as
listed above and can include modifiers, such as if , in , and weight ,
that further restrict or define how the independent variables will be
used in the analysis. if conducts the analysis on only the data as
restricted by this modifier. The in modifier allows you to restrict
variables by a range of observations. Finally, weight can be speci-
fied if the data are weighted in some fashion. With anova there
can be a frequency or analytic weight. In the former, the weighting
indicates the number of duplicate observations, while the latter are
weights indicating the inverse proportion of the weight to the vari-
ance of the form
σ 2
w j
where σ 2 is the variance of an observation and w j is the weight for the
j th observation.
Load the dataset Onion trial 1999.dta into Stata and enter the fol-
lowing command:
anova yieldacre entry rep
This will result in the following output:
Number of obs = 60 R-squared = 0.7882
Root MSE = 101.464 Adj R-squared = 0.6712
Source | Partial SS df MS F Prob > F
-------+----------------------------------------------------
Model | 1455957.67 21 69331.3176 6.73 0.0000
|
entry | 1433977.53 19 75472.5015 7.33 0.0000
rep | 21980.1412 2 10990.0706 1.07 0.3539
|
Residual| 391207.018 38 10294.9215
--------+----------------------------------------------------
Total | 1847164.69 59 31307.8761
This dataset is an onion variety trial that had 20 different varieties
evaluated with three replications arranged in an RCBD. The yieldacre
is the extrapolate yield/acre in 50-lb bags per acre, which was calcu-
lated from the yield variable. The individual plot size or experimental