boot.LR: Parametric bootstrap likelihood-ratio test

In mirt: Multidimensional Item Response Theory

Parametric bootstrap likelihood-ratio test

Description

Given two fitted models, compute a parametric bootstrap test to determine whether the less restrictive models fits significantly better than the more restricted model. Note that this hypothesis test also works when prior parameter distributions are included for either model. Function can be run in parallel after using a suitable mirtCluster definition.

Usage

boot.LR(mod, mod2, R = 1000, verbose = TRUE)

Arguments

mod	an estimated model object, more constrained than mod2
mod2	an estimated model object
R	number of parametric bootstraps to use.
verbose	logical; include additional information in the console?

Value

a p-value evaluating whether the more restrictive model fits significantly worse than the less restrictive model

Author(s)

Phil Chalmers [e-mail]

References

Chalmers, R., P. (2012). mirt: A Multidimensional Item Response Theory Package for the R Environment. Journal of Statistical Software, 48(6), 1-29. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.18637/jss.v048.i06")}

Examples

## Not run: 

# standard
dat <- expand.table(LSAT7)
mod1 <- mirt(dat, 1)
mod2 <- mirt(dat, 1, '3PL')

# standard LR test
anova(mod1, mod2)

# bootstrap LR test (run in parallel to save time)
if(interactive()) mirtCluster()
boot.LR(mod1, mod2, R=200)


## End(Not run)

mirt documentation built on Oct. 17, 2023, 5:06 p.m.