usktest: Unbalanced Scott-Knott (USK)

In felipe179971/USK: Unbalanced Scott-Knott (USK)

Description

usktest is a function used to do the Scott-Knott cluster analysis (1974) for unbalanced designs proposed at 2017 (CONRADO, T. V; FERREIRA, D. F.; SCAPIM, C. A.; MALUF, W. R.). To learn more, see the article.

Usage

usktest(formula, dataset, alpha = 0.05, ANOVA = TRUE)

Arguments

formula	at the moment, this package only does the Unbalanced Scott-Knott for single factor analysis of variance, so your 'formula' must be observation ~ treatment
dataset	an indication to the database being used.
alpha	type I error the researcher can accept (the default is 0.05).
ANOVA	if TRUE (the default), print an ANOVA table.

Value

If ANOVA=TRUE, this function returns the ANOVA table and the Scott-Knott table that are a data.frame including 5 columns:

• 'treatments': the name of the 'treatments' parameter.

• group: which group the treatment was classified.

• mean: average value of 'observation' for each treatment.

• min: lowest value of 'observation' for each treatment.

• max: higher value of 'observation' for each treatment.

Author(s)

Felipe Ferreira [e-mail]

References

CONRADO, T. V; FERREIRA, D. F.; SCAPIM, C. A.; MALUF, W. R. (2017) Adjusting the Scott-Knott cluster analyses for unbalanced designs.

See Also

For graphics: plot_usk or plotly_usk. See also the packages ScottKnott, ScottKnottESD and multcomp.

Examples

set.seed(3)
taus=c(4,4,-4,-4,9,-9)
Tratamento<-as.factor(rep(c(paste("trat",seq(1:length(taus)))),3))
erro<-rnorm(3*length(taus),0,1)
y<-2+taus+erro
y[round(runif(1,min=1,max=length(y)),0)]<-NA
dados<-data.frame(y,Tratamento)

test<-usktest(y~Tratamento,dados)

#ggplot2
plot_usk(test)
#plotly
#plotly_usk(test)

felipe179971/USK documentation built on Dec. 20, 2021, 7:49 a.m.