waller.test: Multiple comparisons, Waller-Duncan

Description

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.

Usage

waller.test(y, trt, DFerror, MSerror, Fc, K = 100, group=TRUE, main = NULL, 
console=FALSE)

Value

statistics: Statistics of the model
parameters: Design parameters
means: Statistical summary of the study variable
comparison: Comparison between treatments
groups: Formation of treatment groups

Arguments

y: model(aov or lm) or answer of the experimental unit
trt: Constant( only y=model) or vector treatment applied to each unit
DFerror: Degrees of freedom
MSerror: Mean Square Error
Fc: F Value
K: K-RATIO
group: TRUE or FALSE
main: Title
console: logical, print output

Author

Felipe de Mendiburu

Details

It is necessary first makes a analysis of variance.

K-RATIO (K): value specifies the Type 1/Type 2 error seriousness ratio for the Waller-Duncan test. Reasonable values for KRATIO are 50, 100, and 500, which roughly correspond for the two-level case to ALPHA levels of 0.1, 0.05, and 0.01. By default, the procedure uses the default value of 100.

if y = model, then to apply the instruction:
waller.test (model, "trt", alpha = 0.05, group = TRUE, main = NULL, console = FALSE)
where the model class is aov or lm.

References

Waller, R. A. and Duncan, D. B. (1969). A Bayes Rule for the Symmetric Multiple Comparison Problem, Journal of the American Statistical Association 64, pages 1484-1504.

Waller, R. A. and Kemp, K. E. (1976) Computations of Bayesian t-Values for Multiple Comparisons, Journal of Statistical Computation and Simulation, 75, pages 169-172.

Steel & Torry & Dickey. Third Edition 1997 Principles and procedures of statistics a biometrical approach

Examples

Run this code

library(agricolae)
data(sweetpotato)
model<-aov(yield~virus, data=sweetpotato)
out <- waller.test(model,"virus", group=TRUE)
#startgraph
oldpar<-par(mfrow=c(2,2))
# variation: SE is error standard
# variation: range is Max - Min
bar.err(out$means,variation="SD",horiz=TRUE,xlim=c(0,45),bar=FALSE,
col=colors()[25],space=2, main="Standard deviation",las=1)
bar.err(out$means,variation="SE",horiz=FALSE,ylim=c(0,45),bar=FALSE,
col=colors()[15],space=2,main="SE",las=1)
bar.err(out$means,variation="range",ylim=c(0,45),bar=FALSE,col="green",
space=3,main="Range = Max - Min",las=1)
bar.group(out$groups,horiz=FALSE,ylim=c(0,45),density=8,col="red", 
main="Groups",las=1)
#endgraph
# Old version HSD.test()
df<-df.residual(model)
MSerror<-deviance(model)/df
Fc<-anova(model)["virus",4]
out <- with(sweetpotato,waller.test(yield, virus, df, MSerror, Fc, group=TRUE))
print(out)
par(oldpar)

Run the code above in your browser using DataLab