Agriculture Reference
In-Depth Information
Again, we will merge this dataset with one we created earlier. Enter
the command
merge 1:1 _n using trttotals.dta
Notice a couple of differences in how the merge command is used
here compared to previously. here are no quotes around the file name
because they are only needed if a file name contains spaces. In addi-
tion, there is no option keepusing() because the entire file on the
disk is being merged into the dataset in memory, not just a particular
variable. Opening the Data Editor will show a dataset of the 16 treat-
ments with the treatment totals as well as the block totals for each
treatment. he _ merge variable can be dropped with
drop _merge
This dataset consists of the totals for each treatment and the block
totals for each treatment. For example, the total for treatment 2 adds
the values (see Chapter 5, TableĀ 5.2)
152 + 155 + 130 + 152 + 205 = 794
For the block totals, each block in which treatment 2 occurs is added.
Treatment 2 occurs in blocks 1, 6, 10, 14, and 18. For block 1, add all
the experimental units in this block (see Chapter 5, TableĀ 5.2):
147 + 152 + 167 + 150 = 616
For all the blocks in which treatment 2 occurs, add the block totals:
616 + 586 + 654 + 724 + 742 = 3,322
At this point, a series of commands are entered to generate several
new variables and local marcos, which will be used to calculate the
adjusted treatment mean square and the adjusted error mean square.
The adjustment factor u , which was previously calculated, is used in
these commands. Because u is calculated from the scalars of the most
recent anova command, you may wish to check if it is still valid. To
do this enter
display `u'
Search WWH ::




Custom Search