brant (version 0.3-0)

brant: Brant Test

Description

The function calculates the brant test by Brant (1990) <doi: 10.2307/2532457> for ordinal logit models to test the parallel regression assumption.

Usage

brant(model,by.var=F)

Arguments

model

the polr-Object generated with polr()

by.var

OPTIONAL if set to true, the tests are made for each variable instead of each coefficient. Default: FALSE.

Value

The output is the brant test, which shows if the parallel assumption holds or not.

Details

The function calculates the brant test for parallel regression assumption. The brant test was published by Brant (1990). The function works with models generated with the function polr() from the package 'MASS'.

References

Brant, R. (1990) Assessing proportionality in the proportional odds model for ordinal logistic regression. Biometrics, 46, 1171--1178.

Examples

# NOT RUN {
data = MASS::survey
data$Smoke = ordered(MASS::survey$Smoke,levels=c("Never","Occas","Regul","Heavy"))
model1 = MASS::polr(Smoke ~ Sex + Height, data=data, Hess=TRUE)
summary(model1)
brant(model1)
# }