StanDINA.run: DCM calibration under the DINA model via Stan

In JihongZ/StanDCM: Diagnostic Classification Model using Stan

Description

StanDINA uses Stan program to calibrate the deterministic inputs, noisy and gate model for dichotomous responses, and its extension.

In addition, users are allowed to specify design matrix and link function for each item, and distinct models may be used in a single test for different items. The attributions can be either dichomous or polytomous.

Usage

StanDINA.run(
  Qmatrix,
  response.matrix,
  script.path = NA,
  save.path = getwd(),
  save.name = "DINA_uninf",
  iter = 1000,
  warmup = 0,
  chain.num = 3,
  init.list = "random",
  control.list = NA
)

Arguments

Qmatrix	the Q-matrix specified for the LCDM
response.matrix	the response matrix
script.path	the path to save the .stan file to somewhere; the default path is getwd()
save.path	the path to save the .stan file to somewhere; the default path is getwd()
save.name	name of the .stan
iter	number of iterations
warmup	number of warmup iterations
chain.num	number of MCMC chains. Default is 3.
init.list	initial variables. Default is random. Other options include cdm
control.list	constrains put on the multigroup DINO model.

Value

StanDINA returens an object of class StanDINA. Methods for StanDINA objects include extract for extract for extracting various components, coef for extracting structural parameters.

Author(s)

Zhehan Jiang, University of Alabama, [e-mail]
Jihong Zhang, University of Iowa, [e-mail]

Examples

## Not run: 
#----------- DINA model-----------#
mod1<-StanDINA.run(Qmatrix, respMatrix, iter=20, init.list='cdm', chain.num = 3)
summary(mod1)

## End(Not run)

JihongZ/StanDCM documentation built on June 27, 2020, 7:51 a.m.