TCClookup: Lookup ability values on the test characteristic curve

In CambridgeAssessmentResearch/unimirt: Applications of Unidimensional IRT using the R Package 'mirt'

Lookup ability values on the test characteristic curve

Description

For unidimensional IRT models this function looks up the ability value associated with each whole number raw score within the test characteristic curve. Where no equivalent ability can be found (e.g. for all item answered correctly) or for raw scores below the total expected by pure guessing values of NA will be returned.

Usage

TCClookup(mirtobj, fulliterative = FALSE, which.items = NULL)

Arguments

mirtobj	An estimated IRT model (of class SingleGroupClass) estimated either using mirt or unimirt.
fulliterative	A logical value denoting whether a full iterative method should be used to identify the appropriate abilities for each score or whether approximation using interpolation between (a thousand) fixed ability points will suffice. The default value is FALSE. Setting this to TRUE makes estimation noticeably slower and tends to make very little difference to the resulting abilities.
which.items	an integer vector indicating which items to include in the expected test score. Default uses all possible items.

Value

A data frame showing the mapping between each possible raw score and the corresponding ability value in the test characteristics curve.

See Also

expected.test

Examples

mirt1=unimirt(mathsdata[,1:20],"3")
#method based on interpolation
TCC1=TCClookup(mirt1)
#method based on full iterative identification of ability points
TCC2=TCClookup(mirt1,fulliterative=TRUE)
#demonstrate that differences are trivial (at least in this example)
plot(TCC1$score,TCC1$TCCabil-TCC2$TCCabil,type='l')
#apply to first five items in test
TCC3=TCClookup(mirt1,which.items=1:5)

CambridgeAssessmentResearch/unimirt documentation built on June 10, 2025, 6:03 a.m.