Agriculture Reference
In-Depth Information
The Residual mean square from the previous ANOVA is used to
calculate the correct F value, so the following commands are entered:
local
z = e(df_1)
local
g1
=
e(ss_1)
/
e(df_1)
/
`x'
display
`g1'
display Ftail
(`z', `y', `g1')
The first command stores the var degrees of freedom, which is 14 in
this case in the macro z. The second command calculates the correct F
value for the first group. The first part of this equation, e(ss_1)/e(df_1),
divides the Partial SS (4.15479519) by the degrees of freedom (14) to
calculate the var mean square (0.296771085), which is then divided by
x, the mean square from the previous ANOVA. The next command
displays the results of this calculation (3.376694). Finally, the last
command calculates the probability associated with the numerator
and denominator degrees of freedom with this F value and displays
the results (0.00025122).
The above shows how the variety mean square is partitioned and
divided by the residual mean square from the overall ANOVA. This
can, however, be easily handled by Stata with the following command:
contrast
var
|
maturity
which results in the following output:
Contrasts of marginal linear predictions
Margins : asbalanced
------------------------------------------------
| df F P>F
-------------+----------------------------------
var|maturity |
1 | 14 3.38 0.0003
2 | 14 2.11 0.0192
3 | 14 4.64 0.0000
Joint | 42 3.37 0.0000
|
Residual | 84
------------------------------------------------
This command must be entered immediately after the estimation
command:
anova
yield rep maturity
/
maturity
#
rep var
|
maturity
Search WWH ::
Custom Search