guesspar {psychotools}

Extract Guessing Parameters of Item Response Models

Description

A class and generic function for representing and extracting the so-called guessing parameters of a given item response model.

Usage

  guesspar(object, ...)
  ## S3 method for class 'raschmodel'
guesspar(object, alias = TRUE, vcov = TRUE, ...)
  ## S3 method for class 'rsmodel'
guesspar(object, alias = TRUE, vcov = TRUE, ...)
  ## S3 method for class 'pcmodel'
guesspar(object, alias = TRUE, vcov = TRUE, ...)
  ## S3 method for class 'nplmodel'
guesspar(object, alias = TRUE, logit = FALSE, vcov = TRUE, ...)
  ## S3 method for class 'gpcmodel'
guesspar(object, alias = TRUE, vcov = TRUE, ...)

Arguments

object

a fitted model object whose guessing parameters should be extracted.

alias

logical. If TRUE (the default), the aliased parameters are included in the return vector (and in the variance-covariance matrix if vcov = TRUE). If FALSE, these parameters are removed. For raschmodels, rsmodels, pcmodels and gpcmodels, where all guessing parameters are fixed to 0, this means that an empty numeric vector and an empty variance-covariace matrix is returned if alias is FALSE.

logit

logical. If a nplmodel of type "3PL" or "4PL" model has been fit, the guessing parameters were estimated on the logit scale. If logit = FALSE, these estimates and the variance-covariance (if requested) are retransformed using the logistic function and the delta method.

vcov

logical. If TRUE (the default), the variance-covariance matrix of the guessing parameters is attached as attribute vcov.

...

further arguments which are currently not used.

Details

guesspar is both, a class to represent guessing parameters of item response models as well as a generic function. The generic function can be used to extract the guessing parameters of a given item response model.

For objects of class guesspar, several methods to standard generic functions exist: print, coef, vcov. coef and vcov can be used to extract the guessing parameters and their variance-covariance matrix without additional attributes.

Value

A named vector with guessing parameters of class guesspar and additional attributes model (the model name), alias (either TRUE or a named numeric vector with the aliased parameters not included in the return value), logit (indicating whether the estimates are on the logit scale or not), and vcov (the estimated and adjusted variance-covariance matrix).

See Also

personpar, itempar, threshpar, discrpar, upperpar

Examples

if(requireNamespace("mirt")) {

o <- options(digits = 3)

## load simulated data
data("Sim3PL", package = "psychotools")

## fit 2PL to data simulated under the 3PL
twoplmod <- nplmodel(Sim3PL$resp)

## extract the guessing parameters (all fixed at 0)
gp1 <- guesspar(twoplmod)

## fit 3PL to data simulated under the 3PL
threeplmod <- nplmodel(Sim3PL$resp, type = "3PL")

## extract the guessing parameters
gp2 <- guesspar(threeplmod)

## extract the standard errors
sqrt(diag(vcov(gp2)))

## extract the guessing parameters on the logit scale
gp2_logit <- guesspar(threeplmod, logit = TRUE)

## along with the delta transformed standard errors
sqrt(diag(vcov(gp2_logit)))

options(digits = o$digits)
}