Estimate IRT scale parameters using the robust DIF procedure.

Estimation can be performed using iteratively re-weighted least squares (IRLS) or Newton-Raphson (NR). Currently, only IRLS is implemented. If starting.value = "all", three starting values are computed: the median of y_fun, the least trimmed squares estimate of location for y_fun with 50-percent trim rate, and the minimum of rho_grid. The estimate is computed from each starting value, and the solution with the lowest value of the bi-square objective function is returned. If there are multiple solutions, they are stored other.solutions.

Usage

rdif(
  mle,
  fun = "d_fun3",
  alpha = 0.05,
  starting.value = "all",
  tol = 1e-07,
  maxit = 100,
  method = "irls"
)

Arguments

mle

the output of get_model_parms

fun

one of c("a_fun1", "a_fun2", "d_fun1", "d_fun2", "d_fun3"). See description for details.

alpha

the desired false positive rate for flagging items with DIF.

starting.value

one of c("med", "lts", "min_rho", "all") or a numerical value to be used as the starting value. See description for details.

tol

convergence criterion for comparing subsequent values of estimate

maxit

maximum number of iterations

method

one of c("irls", "newton"). Currently, only IRLS is implemented.

Value

An rdif object.

Details

Implements M-estimation of an IRT scale parameter using the bi-square loss function. Also returns the bi-square weights for each item.

Examples

# Item intercepts, using the built-in example dataset "rdif.eg"
rdif(mle = rdif.eg, fun = "d_fun3")
#> Error: object 'rdif.eg' not found

# Item slopes
rdif(mle = rdif.eg, fun = "a_fun1")
#> Error: object 'rdif.eg' not found