rgpcm: Simulate Data under a Generalized Partial Credit Model

In psychotools: Psychometric Modeling Infrastructure

Simulate Data under a Generalized Partial Credit Model

Description

rgpcm simulates IRT data under a generalized partial credit model.

Usage

rgpcm(theta, a, b, nullcats = FALSE, return_setting = TRUE)

Arguments

theta	numeric vector of person parameters. Can also be a list, then a list of length length(theta) is returned, containing multiple simulated data sets.
a	list of numerics of item discrimination parameters.
b	list of numeric vectors of item threshold parameters.
nullcats	logical. Should null categories be allowed?
return_setting	logical. Should a list containing slots of "a", "b", and "theta", as well as the simulated data matrix "data" be returned (default) or only the simulated data matrix?

Value

rgpcm returns either a list of the following components:

a	list of numerics of item discrimination parameters used,
b	list of numeric vectors of item threshold parameters used,
theta	numeric vector of person parameters used,
data	numeric matrix containing the simulated data,

or (if return_setting = FALSE) only the numeric matrix containing the simulated data.

See Also

rpcm, rrsm, rpl, rrm

Examples

set.seed(1)
## item responses under a GPCM (generalized partial credit model) from
## 6 persons with three different person parameters
## 8 items with different combinations of two or three threshold parameters
## and corresponding discrimination parameters
ppar <- rep(-1:1, each = 2)
tpar <- rep(list(-2:0, -1:1, 0:1, 0:2), each = 2)
dpar <- rep(list(1, 2), each = 4)
sim <- rgpcm(theta = ppar, a = dpar, b = tpar)

## simulated item response data along with setting parameters
sim

## print and plot corresponding item response object
iresp <- itemresp(sim$data)
iresp
plot(iresp)

psychotools documentation built on May 29, 2024, 8:12 a.m.