sep.pars-class {plink}R Documentation

Class "sep.pars"

Description

The formal S4 class for sep.pars. This class stores a set of separated item parameters and characteristics of these parameters.

Details

a will be an n x 1 matrix for n items if there are no nomial responce model or multiple-choice model items. Otherwise, if nrm or mcm items are included, a will be an n x m matrix with m equal to the maximum number of response categories across items. If nrm or mcm items are included, the discrimination parameters for the dictomous response models, the graded response model, partial credit model and generalized partial credit model are listed in the first column will all other columns filled with NAs.

b is an n x m matrix of difficulty, threshold, step, or category parameters (depending on the corresponding model) for n items with m equal to the maximum number of b parameters across all items. For items with b less than m, the row is right-filled with NAs.

c will be an n x 1 matrix for n items if there are no multiple-choice model items. Otherwise, if mcm items are included, c will be an n x m matrix with m equal to the maximum number of response categories across items minus one. If mcm items are included, the lower asmptote parameters for the dictomous response models, are listed in the first column will all other columns filled with NAs. The c values for the 1PL and 2PL equal zero.

Objects from the Class

Objects can be created by calls of the form new("sep.pars", ...), but this is not encouraged. Use the function sep.pars instead.

Slots

a:
matrix of discrimination parameters
b:
matrix of difficulty, threshold, step, and category parameters (depending on the associated IRT model)
c:
matrix of lower asymptote parameters and category proportions for the dichotomous response models and multiple-choice model respectively. c is equal to NA for each item for all other models.
cat:
vector identifying the number of categories associated with each item. All dichotomous items will have cat values equal to 2. Graded response model and partial credit/generalized partial credit model items will have cat values equal to the number of step/threshold parameters plus one. Nominal response model items will have cat values equal to the number of categories, and multiple-choice model items will have cat values equal to the number of categories plus one (the 'do not know' category).
n:
vector identifying the total number of items, the total number of dichotomous and polytomous items, and the number of items associated with each polytomous model.
mod.lab:
character vector of labels for the model(s).

loc.out:
logical value. If TRUE, the step and/or threshold parameters in slot b for the graded response model and generalized partial credit model include a location parameter.
model:
character vector identifying all the models associated with the corresponding set of item parameters. The only acceptable models are drm, gpcm, grm, mcm, and nrm (see class "poly.mod" for more information).
items:
list with the same length as model, where each element identifies the items(rows) in the corresponding set of item parameters associated with the model(s) identified in model.

Extends

Class "poly.mod", directly.
Class "list.poly", by class "poly.mod", distance 2.

Author(s)

Jonathan P. Weeks weeksjp@gmail.com

See Also

sep.pars, irt.pars


[Package plink version 0.1-1 Index]