dgp_uni: Data generating process for unidimensional rasch model

Description Usage Arguments Value References Examples

View source: R/uniDimDGP.R

Description

Data generating process for unidimensional rasch model

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
dgp_uni(
  nobs,
  tlength,
  DIFpercent,
  DIFpattern = "balanced",
  DIFeffect = "constant",
  DIFamount = 0.6,
  ability = TRUE,
  sigmaable = c(1, 1),
  itemref = c(-2.522, -1.902, -1.351, -1.092, -0.234, -0.317, 0.037, 0.268, -0.571,
    0.317, 0.295, 0.778, 1.514, 1.744, 1.951, -1.152, -0.526, 1.104, 0.961, 1.314,
    -2.198, -1.621, -0.761, -1.179, -0.61, -0.291, 0.067, 0.706, -2.713, 0.213, 0.116,
    0.273, 0.84, 0.745, 1.485, -1.208, 0.189, 0.345, 0.962, 1.592)
)

Arguments

nobs

number of observations per group

tlength

interger > 0, test length (number of items)

DIFpercent

percentage of DIF items in the test

DIFpattern

"balanced": DIF balanced over groups "favorref","favorfoc": all DIF items favor one group

DIFeffect

data generating process for DIF effect:

  • normal: item parameter differences are drawn at random from a normal distribution with mean DIFamount and sd = 0.1, like in Wang et al. (2012)

  • uniform: item parameter differences are drawn at random from the vector [DIFamount-0.4,DIFamount-0.2,DIFamount,DIFamount+0.2,DIFamount+0.4]

  • constant: item parameter differences are defined as DIFamount for all items

DIFamount

magnitude of DIF

ability

should the groups differ in mean ability? (default is TRUE)

sigmaable

positive numeric vector of length two, standard deviations for person parameter distributions in the two groups (default is c(1,1))

itemref

numeric vector of length tlength (if shorter, then sampling with replacement is used), item difficulty parameter for reference group like in Wang et al. (2012)

Value

list containing:

References

Examples

1
# For examples, see ?getData.

lucasmanuelkohler/anchorpoint documentation built on April 16, 2021, 6:41 a.m.