SL {MiscPsycho}R Documentation

Stocking-Lord Procedure

Description

This function estimates the A and B parameters (interept and slope) such that two tests with common items can be linked. Common items must be multiple choice items based on the 3-parameter logistic model.

Usage

SL(params1, params2, con = 0.001, ...)

Arguments

params1 item parameters of the common items used for linking organized as a list of lists for test A. See irt.ability for more details.
params2 item parameters of the common items used for linking organized as a list of lists for test B. See irt.ability for more details.
con Criterion for convergence
... a list of control parameters. These are actually passed from the function SLderivs,
D
a constant usually fixed at 1.7 to bring the logistic function into coincidence with the probit.
mu
mean of the normal prior distribution.
sigma
standard deviation of the normal prior distribution.
Q
number of quadrature points used in the Gauss-Hermite approximation.

Value

A list with two values for the linking constants

A The Stocking-Lord intercept
B The Stocking-Lord slope

Author(s)

Harold Doran

References

Kolen, M.J., Brennan, R.L. (2004). Test equating, scaling, and linking. Second Edition. Springer.

Examples

## This is 3PL and are values from page 171 of Kolen and Brennan
params1 <- list("3pl" = list(a = c(.4,1.7,1.2), b = c(-1.1, .9,2.2), c = c(.1,.2,.1)), "gpcm" = NULL)
params2 <- list("3pl" = list(a = c(.5,1.6,1), b = c(-1.5, .5,2), c = c(.1,.2,.1)), "gpcm" = NULL)
SL(params1, params2, control=list(Q=30, mu=0, sigma=1))

[Package MiscPsycho version 1.3 Index]