LScottKnott {laercio} | R Documentation |
Scott-Knott test to group means. This function was tested and it worked correctly for completely random design and randomized complete block design.
LScottKnott(anova, which, conf.level=0.95)
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. |
Laercio Junio da Silva laerciojunio@yahoo.com.br
GATES, C.E.; BILBRO, J.D. Illustration of a Cluster Analysis Method for Mean Separation. Agronomy Journal, Vol.70, May-June 1978.
#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")