Agriculture Reference
In-Depth Information
This will result in the following output:
Predictive margins Number of obs = 12
Expression : Linear prediction, predict()
----------------------------------------------------------------
| Delta-method
| Margin Std. Err. z P>|z| [95% Conf. Interval]
------+---------------------------------------------------------
temp |
10 | .0235833 .0450639 0.52 0.601 -.0647404 .111907
20 | .4334583 .0450639 9.62 0.000 .3451346 .5217821
30 | .4928333 .0450639 10.94 0.000 .4045096 .581157
40 | .0724583 .0450639 1.61 0.108 -.0158654 .1607821
----------------------------------------------------------------
The first column labeled Margin contains the marginal or least squares
means. To compare it to the arithmetic means, enter
mean
root, over(temp)
which results in
Mean estimation Number of obs = 12
10: temp = 10
20: temp = 20
30: temp = 30
40: temp = 40
--------------------------------------------------------------
Over | Mean Std. Err. [95% Conf. Interval]
-------------+------------------------------------------------
root |
10 | .0023333 .001453 -.0008646 .0055313
20 | .4366667 .031798 .3666798 .5066535
30 | .5033333 .0768837 .3341133 .6725533
40 | .08 .0152753 .0463794 .1136206
--------------------------------------------------------------
Although there is not a great deal of difference between the two sets
of means, there are differences particularly for the 10°C treatment.
Finally, if you open the Data Editor window and look at this data-
set, you will notice missing values for dependent variable, root. These
observations do not have to be part of the dataset for the calculations
Search WWH ::
Custom Search