Agriculture Reference
In-Depth Information
blocks in the field that are used to account for potential differences
that may occur because of location. In this particular case, the LS
accounts for two additional sources of variation, both row and col-
umn position. From a research perspective, the fertilizer treatments
do have an effect on bulb pyruvate with a significance of Prob > F of
0.0037, which would be of primary interest.
The column variable is significant, but the row variable is not. It is
possible to calculate the relative efficiency of the LS design compared
to a CRD as well as the RCBD. These formulas include
(
)
EE t
++−
1
E
.
(
)
=
r
c
e
RE CRD
(
)
t
+
1
E
e
+
( )
()()
Et
1
E
(
)
=
r
e
RE RCBrow
..
,
tE
e
+
( )
()()
Et
1
E
(
)
=
c
e
RE RCTcolumn
..
,
tE
e
There is also a correction factor for the LS design that should be
used if the error degrees of freedom is below 20. In this case it is at 6.
The correction factor is
(
)
(
)
+
(
)
+
2
t
−
1
t
−
2
1
t
−
1
3
k
=
(
)
(
)
+
(
)
+
2
t
−
1
t
−
2
3
t
−
1
1
The following listing uses the scalars from the previous estimation
and calculates the relative efficiencies of the LS design compared to
CRD and RCBD designs. Entering this sequence of commands can
be tedious and error prone, so I have already done it for you. Open
the Do-File LS Efficiency.do in the Do-File folder. Once open in a
Do-file Editor, click the Do icon in the upper right-hand corner of
the editor. This will display the four calculated values: k, RE, RER,
and REC, which represent the correction factor, relative efficiency,
relative efficiency of the rows, and relative efficiency of the columns,
respectively. The code segment is an example of a Do-File that will be
Search WWH ::
Custom Search