Multiple comparisons, "Least significant difference" and Adjust P-values

Source: R/LSD.test.R

Multiple comparisons of treatments by means of LSD and a grouping of treatments. The level by alpha default is 0.05. Returns p-values adjusted using one of several methods

LSD.test(
  y,
  trt,
  DFerror,
  MSerror,
  alpha = 0.05,
  p.adj = c("none", "holm", "hommel", "hochberg", "bonferroni", "BH", "BY", "fdr"),
  group = TRUE,
  main = NULL,
  console = FALSE
)

Arguments

y

model(aov or lm) or answer of the experimental unit
trt

Constant( only y=model) or vector treatment applied to each experimental unit
DFerror

Degrees of freedom of the experimental error
MSerror

Means square error of the experimental
alpha

Level of risk for the test
p.adj

Method for adjusting p values (see p.adjust)
group

TRUE or FALSE
main

title of the study
console

logical, print output

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

Details

For equal or different repetition.
For the adjustment methods, see the function p.adjusted.
p-adj ="none" is t-student.

It is necessary first makes a analysis of variance.
if model=y, then to apply the instruction:
LSD.test(model, "trt", alpha = 0.05, p.adj=c("none","holm","hommel", "hochberg", "bonferroni", "BH", "BY", "fdr"), group=TRUE, main = NULL,console=FALSE)
where the model class is aov or lm.

References

Steel, R.; Torri,J; Dickey, D.(1997) Principles and Procedures of Statistics A Biometrical Approach. pp178.

See also

BIB.test, DAU.test, duncan.test, durbin.test, friedman, HSD.test, kruskal, Median.test, PBIB.test, REGW.test, scheffe.test, SNK.test, waerden.test, waller.test, plot.group

Examples


library(agricolae)
data(sweetpotato)
model<-aov(yield~virus, data=sweetpotato)
out <- LSD.test(model,"virus", p.adj="bonferroni")
#stargraph
# Variation range: max and min
plot(out)
#> Warning: NAs introduced by coercion
#endgraph
# Old version LSD.test()
df<-df.residual(model)
MSerror<-deviance(model)/df
out <- with(sweetpotato,LSD.test(yield,virus,df,MSerror))
#stargraph
# Variation interquartil range: Q75 and Q25
plot(out,variation="IQR")
#> Warning: NAs introduced by coercion
#endgraph
out<-LSD.test(model,"virus",p.adj="hommel",console=TRUE)
#> 
#> Study: model ~ "virus"
#> 
#> LSD t Test for yield 
#> P value adjustment method: hommel 
#> 
#> Mean Square Error:  22.48917 
#> 
#> virus,  means and individual ( 95 %) CI
#> 
#>       yield      std r       LCL      UCL  Min  Max
#> cc 24.40000 3.609709 3 18.086268 30.71373 21.7 28.5
#> fc 12.86667 2.159475 3  6.552935 19.18040 10.6 14.9
#> ff 36.33333 7.333030 3 30.019601 42.64707 28.0 41.8
#> oo 36.90000 4.300000 3 30.586268 43.21373 32.1 40.4
#> 
#> Alpha: 0.05 ; DF Error: 8
#> Critical Value of t: 2.306004 
#> 
#> Minimum Significant Difference: 8.928965 
#> 
#> Treatments with the same letter are not significantly different.
#> 
#>       yield groups
#> oo 36.90000      a
#> ff 36.33333      a
#> cc 24.40000      b
#> fc 12.86667      c
plot(out,variation="SD") # variation standard deviation
#> Warning: NAs introduced by coercion