genResponses: genResponses

In conquestr: An R Package to Extend 'ACER ConQuest'

genResponses

Description

Generates response vectors for n cases to i items given known item parameters, person abilities, and (optionally) other inputs.

Usage

genResponses(abilities, itemParams, BMatrix = 1, mcarP = 0, perturbP = NULL)

Arguments

abilities	A person by latent-dimension matrix of abilities. One column per dimension.
itemParams	A list of item params of the structure used in simplef (a matrix of k categories by four (category score, delta dot, tau, discrimintation)). See conquestr::makeItemList for a helper to generate this list.
BMatrix	A simplified B-matrix mapping dimensions (columns) to items (rows). Or the integer "1" if items are dichotomous and ability is uni-dimensional.
mcarP	A double indicating the proportion of missing data under the MCAR assumption.
perturbP	A list, where each element of the list contains a data frame referring to an item. Each data frame is either a 1 * 4 data frame describing the general perturbation to apply to the probabilities for that item (currently developed) or a pair of vectors mapping theta to probabilities (not developed). In the first case - the data frame is in this order: "item": item number (int), "type": the type of perturbation to apply (string, "flat", or "steep"), "pivot": probability pivot point around which the perturbation is applied (double in0 < x < 1), "factor": magnitude of the perturbation (double - when in 0 < x < 100 probs remain positively correlated with theta (0 = no perturbation, 100 = maximum perturbation) when in100 < x < Inf probs are negatively correlated with theta).

Value

A matrix, n cases by i items, of scored item responses.

See Also

simplef(), browseVignettes("conquestr")

Examples

  myItem <- matrix(c(0, 0, 0, 0, 1, 1, 0, 1), ncol = 4, byrow = TRUE)
  myItems <- list(myItem, myItem)
  myItems[[2]][2, 2] <- -1 # make the second item delta equal to -1
  myResponses <- genResponses(rnorm(100), myItems)

conquestr documentation built on March 31, 2023, 6:02 p.m.