VCA (version 1.5.1)

test.lsmeans: Perform t-Tests for Linear Contrasts on LS Means

Description

Perform custom hypothesis tests on Least Squares Means (LS Means) of fixed effect.

Usage

test.lsmeans(
  obj,
  L,
  ddfm = c("contain", "residual", "satterthwaite"),
  quiet = FALSE
)

Arguments

obj

(VCA) object

L

(matrix) specifying one or multiple custom hypothesis tests as linear contrasts of LS Means. Which LS Means have to be used is inferred from the column names of matrix \(L\), which need to be in line with the naming of LS Means in function lsmeans.

ddfm

(character) string specifying the method used for computing the denominator degrees of freedom of t-tests of LS Means. Available methods are "contain", "residual", and "satterthwaite".

quiet

(logical) TRUE = will suppress any warning, which will be issued otherwise

Author

Andre Schuetzenmeister

Details

This function is similar to function test.fixef and represents a convenient way of specifying linear contrasts of LS Means.

See Also

test.fixef, lsmeans

Examples

if (FALSE) {
data(dataEP05A2_2)
ub.dat <- dataEP05A2_2[-c(11,12,23,32,40,41,42),]
fit1 <- anovaMM(y~day/(run), ub.dat)
fit2 <- remlMM(y~day/(run), ub.dat)
lsm1 <- lsmeans(fit1)
lsm2 <- lsmeans(fit2)
lsm1
lsm2

lc.mat <- getL(fit1, c("day1-day2", "day3-day6"), "lsm")
lc.mat[1,c(1,2)] <- c(1,-1)
lc.mat[2,c(3,6)] <- c(1,-1)
lc.mat
test.lsmeans(fit1, lc.mat) 
test.lsmeans(fit2, lc.mat)

# fit mixed model from the 'nlme' package

library(nlme)
data(Orthodont)
fit.lme <- lme(distance~Sex*I(age-11), random=~I(age-11)|Subject, Orthodont) 

# re-organize data for using 'anovaMM'
Ortho <- Orthodont
Ortho$age2 <- Ortho$age - 11
Ortho$Subject <- factor(as.character(Ortho$Subject))

# model without intercept
fit.anovaMM <- anovaMM(distance~Sex+Sex:age2+(Subject)+(Subject):age2-1, Ortho)
fit.remlMM1 <- remlMM( distance~Sex+Sex:age2+(Subject)+(Subject):age2-1, Ortho)
fit.remlMM2 <- remlMM( distance~Sex+Sex:age2+(Subject)+(Subject):age2-1, Ortho, cov=FALSE)
lsm0 <- lsmeans(fit.anovaMM)
lsm1 <- lsmeans(fit.remlMM1)
lsm2 <- lsmeans(fit.remlMM2)
lsm0
lsm1
lsm2

lc.mat <- matrix(c(1,-1), nrow=1, dimnames=list("int.Male-int.Female", c("SexMale", "SexFemale")))
lc.mat
test.lsmeans(fit.anovaMM, lc.mat)	
test.lsmeans(fit.remlMM1, lc.mat)
test.lsmeans(fit.remlMM2, lc.mat)
}