Task View ExperimentalDesign
agricolae comprises the functionality of statistical analysis into experimental designs applied specially for field experiments in agriculture and plant breeding: Lattice, factorial, complete and incomplete block, Latin Square, Greaco, Alpha designs, Cyclic designs, Split and strip plot design, augmented block design, comparison of multi-location trials, comparison between treatments, re-sampling, simulation, biodiversity indexes and consensus cluster.
The Post Hoc test (Multiple comparisons)
Parametric methods:
- Fisher: Least significant difference (LSD)
and Adjust P-values (Bonferroni,
- Tukey: Honestly significant difference (HSD)
- Duncan: Minimum significante difference (MSD)
- Waller-Duncan:
The Waller-Duncan k-ratio t test is performed on all main effect means in the
MEANS statement. See the K-RATIO option for information on controlling details
of the test.
- Scheffe: Scheffe 1959, method is very general in that all
possible contrasts can be tested for significance and confidence intervals can
be constructed for the corresponding linear. The test is conservative.
- Student Newman Keul
(SNK) is derived from Tukey, but it is less conservative (finds more
differences). Tukey controls the error for all comparisons, where SNK only
controls for comparisons under consideration. The level by alpha default is
0.05.
- Ryan, Einot and Gabriel and Welsch multiple range test (REGW).
Multiple range tests for all pairwise comparisons, to obtain a confident
inequalities multiple range tests.
Non-parametric methods:
- Kruskal-Wallis.
It makes the multiple comparison with Kruskal-Wallis. The alpha parameter by
default is 0.05. Post hoc test is using the criterium Fisher's least significant
difference. The adjustment methods include the Bonferroni correction and others.
- Friedman test. The data consists of b mutually independent
blocks of k random variables. The data conforme a table b by k. A ranking is
established in each block, ranking values are used in the comparison of the k
variables.
- Median test
for several independent samples. The median test is designed to examine whether
several samples came from populations having the same median.
- The Van Der Waerden (Normal Scores).
The data consist of k samples of possibly unequal sample size. The post hoc test
uses Fisher's minimum criterion difference.
Version 1.0: December 09, 2006
Version 1.2-7 August 31, 2017
August 2017.