sim_model function

Simulate an Item Response Model

Simulates an item response model given a fitted object or input of item response probabilities and skill class probabilities.

sim_model(object=NULL, irfprob=NULL, theta_index=NULL, prob.theta=NULL,
      data=NULL, N_sim=NULL )

Arguments

• object: Fitted object for which the methods IRT.posterior, and IRT.data are defined.
• irfprob: Array of item response function values (items $\times$ categories $\times$ skill classes)
• theta_index: Skill class index for sampling
• prob.theta: Skill class probabilities
• data: Original dataset, only relevant for simulating item response pattern with missing values
• N_sim: Number of subjects to be simulated

Returns

List containing elements - dat: Simulated item responses

• theta: Simulated skill classes

• theta_index: Corresponding indices to theta

Examples

## Not run:

#############################################################################
# EXAMPLE 1: GDINA model simulation
#############################################################################

data(sim.dina, package="CDM")
data(sim.qmatrix, package="CDM")
dat <- sim.dina
Q <- sim.qmatrix

# fit DINA model
mod <- CDM::gdina( dat, q.matrix=Q, rule="DINA")
summary(mod)

#** simulate new item responses (N equals observed sample size)
dat1 <- CDM::sim_model(mod)

#*** simulate item responses for N=2000 subjects
dat2 <- CDM::sim_model(mod, N_sim=2000)
str(dat2)

#*** simulate item responses based on input item response probabilities
#*** and theta_index
irfprob <- CDM::IRT.irfprob(mod)
prob.theta <- attr(irfprob, "prob.theta")
TP <- length(prob.theta)
theta_index <- sample(1:TP, size=1000, prob=prob.theta, replace=TRUE )
#-- simulate
dat3 <- CDM::sim_model(irfprob=irfprob, theta_index=theta_index)
str(dat3)
## End(Not run)