Agriculture Reference
In-Depth Information
Number of obs = 80 R-squared = 0.7531
Root MSE = 17.9712 Adj R-squared = 0.5665
Source | Seq. SS df MS F Prob > F
---------+----------------------------------------------------
Model | 44322.2375 34 1303.59522 4.04 0.0000
|
rep | 5946.05 4 1486.5125 4.60 0.0034
trt | 26994.35 15 1799.62333 5.57 0.0000
block | 11381.8375 15 758.789167 2.35 0.0138
|
Residual | 14533.3125 45 322.9625
---------+----------------------------------------------------
Total | 58855.55 79 745.006962
Balanced Lattice Design with Adjustments
Treatment (adj.) MS: 1600.116667
Effective error (residual) MS: 369.3375921
Computed F: 4.332395892
Prob > F: 0.0001
Coefficient of Variation: 11.2%
Relative Efficiency over an RCB: 17%
The adjustment did not result in a significantly different result from
the original analysis, but this will not always be the case. In addition,
the coefficient of variation (CV) and the relative efficiency compared
to the RCBD are calculated. The specifics of the calculations of this
Do-File are presented in the Appendix. Gomez and Gomez (1984)
have a good presentation of this analysis.
Group Balanced Block Design
In the previous section, to help control variability, a new factor was
introduced, the block, which helps control variability in the experi-
ment due to field position. The group balanced block design attempts
to control variability by identifying a factor associated with the treat-
ments themselves. This design may be used with large variety trials
where, for example, maturity class or growth habit may be distinctive
among the varieties.
The design is arranged much like an RCBD with the difference
that the treatments are randomized within groups in each replication.
So, for example, in a trial of 45 varieties with 3 groups of 15 variet-
ies of different maturity, the varieties would be randomized within
each group within a replication. Because of the way the experiment is
Search WWH ::
Custom Search