LScottKnott: Scott-Knott test

In laercio: Duncan test, Tukey test and Scott-Knott test.

Description

Scott-Knott test to group means. This function was tested and it worked correctly for completely random design and randomized complete block design.

Usage

LScottKnott(anova, which, conf.level=0.95)

Arguments

anova	A aov fitted model object.
which	The factor in the fitted model for be used in the test.
conf.level	A numeric value between zero and one. Default is 0.95.

Author(s)

Laercio Junio da Silva [e-mail]

References

GATES, C.E.; BILBRO, J.D. Illustration of a Cluster Analysis Method for Mean Separation. Agronomy Journal, Vol.70, May-June 1978.

Examples

#completely random design
require(laercio)
ldata$trat <- factor(ldata$trat)
anv <- aov(resp~trat,ldata)
anova(anv)
LScottKnott(anv,"trat")

#randomized complete block design
require(laercio)
ldata$trat <- factor(ldata$trat)
ldata$block <- factor(ldata$block)
anv <- aov(resp~trat+block,ldata)
anova(anv)
LScottKnott(anv,"trat")
LScottKnott(anv,"block")

laercio documentation built on May 2, 2019, 6:13 a.m.