Jackknife Equating Error in Generalized Logistic Rasch Model
This function estimates the linking error in linking based on Jackknife (Monseur & Berezner, 2007).
equating.rasch.jackknife(pars.data, display=TRUE, se.linkerror=FALSE, alpha1=0, alpha2=0)
pars.data |
Data frame with four columns: jackknife unit (1st column), item parameter study 1 (2nd column), item parameter study 2 (3rd column), item (4th column) |
display |
Display progress? |
se.linkerror |
Compute standard error of the linking error |
alpha1 |
Fixed α_1 parameter in the generalized item response model |
alpha2 |
Fixed α_2 parameter in the generalized item response model |
A list with following entries:
pars.data |
Used item parameters |
itemunits |
Used units for jackknife |
descriptives |
Descriptives for Jackknife.
|
Monseur, C., & Berezner, A. (2007). The computation of equating errors in international surveys in education. Journal of Applied Measurement, 8, 323-335.
For more details on linking methods see equating.rasch.
#############################################################################
# EXAMPLE 1: Linking errors PISA study
#############################################################################
data(data.pisaPars)
pars <- data.pisaPars
# Linking error: Jackknife unit is the testlet
vars <- c("testlet","study1","study2","item")
res1 <- sirt::equating.rasch.jackknife(pars[, vars])
res1$descriptives
## N.items N.units shift SD linkerror.jackknife SE.SD.jackknife
## 1 25 8 0.09292838 0.1487387 0.04491197 0.03466309
# Linking error: Jackknife unit is the item
res2 <- sirt::equating.rasch.jackknife(pars[, vars ] )
res2$descriptives
## N.items N.units shift SD linkerror.jackknife SE.SD.jackknife
## 1 25 25 0.09292838 0.1487387 0.02682839 0.02533327Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.