iteminfo: Function to calculate item information

In mirt: Multidimensional Item Response Theory

Function to calculate item information

Description

Given an internal mirt item object extracted by using extract.item, compute the item information.

Usage

iteminfo(x, Theta, degrees = NULL, total.info = TRUE, multidim_matrix = FALSE)

Arguments

x	an extracted internal mirt object containing item information (see extract.item)
Theta	a vector (unidimensional) or matrix (multidimensional) of latent trait values
degrees	a vector of angles in degrees that are between 0 and 90. Only applicable when the input object is multidimensional
total.info	logical; return the total information curve for the item? If FALSE, information curves for each category are returned as a matrix
multidim_matrix	logical; compute the information matrix for each row in Theta? If Theta contains more than 1 row then a list of matrices will be returned, otherwise if Theta has exactly one row then a matrix will be returned

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")}

See Also

extract.item

Examples

mod <- mirt(Science, 1)
extr.2 <- extract.item(mod, 2)
Theta <- matrix(seq(-4,4, by = .1))
info.2 <- iteminfo(extr.2, Theta)

#do something with the info?
plot(Theta, info.2, type = 'l', main = 'Item information')

## Not run: 

#category information curves
cat.info <- iteminfo(extr.2, Theta, total.info = FALSE)
plot(Theta, cat.info[,1], type = 'l', ylim = c(0, max(cat.info)),
     ylab = 'info', main = 'Category information')
for(i in 2:ncol(cat.info))
   lines(Theta, cat.info[,i], col = i)

## Customized test information plot
T1 <- T2 <- 0
dat <- expand.table(LSAT7)
mod1 <- mirt(dat, 1)
mod2 <- mirt(dat, 1, 'Rasch')
for(i in 1:5){
  T1 <- T1 + iteminfo(extract.item(mod1, i), Theta)
  T2 <- T2 + iteminfo(extract.item(mod2, i), Theta)
}
plot(Theta, T2/T1, type = 'l', ylab = 'Relative Test Information', las = 1)
lines(Theta, T1/T1, col = 'red')

# multidimensional
mod <- mirt(dat, 2, TOL=1e-2)
ii <- extract.item(mod, 1)
Theta <- as.matrix(expand.grid(-4:4, -4:4))

iteminfo(ii, Theta, degrees=c(45,45)) # equal angle
iteminfo(ii, Theta, degrees=c(90,0)) # first dimension only

# information matrices
iteminfo(ii, Theta, multidim_matrix = TRUE)
iteminfo(ii, Theta[1, , drop=FALSE], multidim_matrix = TRUE)


## End(Not run)

mirt documentation built on Sept. 11, 2024, 7:14 p.m.