GVEM Algorithm for the Graded Response Model

Usage

MGRM_gvem(
  data,
  model = matrix(1, ncol(data)),
  method = "GVEM",
  iter = 200,
  tol = 1e-04,
  S = 10,
  M = 10,
  MinDim = 0,
  MaxDim = 0,
  verbose = FALSE,
  EFA = FALSE
)

Arguments

data: An $N\times J$ matrix of item responses where 0 is the minimal partial credit score (missing responses should be coded as NA)
model: A $J\times K$ matrix of loading indicators (K is the Number of latent dimension)(all items load on the only dimension by default)
iter: Maximum number of iterations
tol: Termination criterion on numerical accuracy
S: Sample size for approximating the expected lower bound ('IWGVEM' only)
M: Sample size for approximating a tighter lower bound ('IWGVEM' only)
MinDim: Minimum num of possible dimensions ('EFA' only)
MaxDim: Maximum num of possible dimensions ('EFA' only)
verbose: Whether to show the progress
EFA: Whether to run EFA or CFA
criterion: Information criterion for model selection, one of 'GIC' (recommended), 'BIC', or 'AIC'
c: Constant for computing GIC

Value

An object of class vemirt_DIF, which is a list containing the following elements:

...$SIGMA: Person-level posterior covariance matrices
...$MU: Person-level posterior mean vectors
...$Sigma: Group-level covariance matrices
...$Mu: Group-level mean vectors
...$ksi1: Variational parameter 1
...$ksi2: Variational parameter 2
...$dim: Num of dimension between latent variables
...$a: Slopes
...$b: Intercepts
...$n2vlb: Bayesian Information Criterion: -2*ll+l0*log(N)
iter: Number(s) of iterations for initialization

Author

Yijun Cheng <chengxb@uw.edu>

Examples

if (FALSE) { # \dontrun{
with(MGRM_data, MGRM_gvem(data, method = "IWGVEM", model, EFA = FALSE))} # }