1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
|
\name{discrpar}
\alias{discrpar}
\alias{discrpar.raschmodel}
\alias{discrpar.rsmodel}
\alias{discrpar.pcmodel}
\alias{discrpar.plmodel}
\alias{discrpar.gpcmodel}
\alias{coef.discrpar}
\alias{print.discrpar}
\alias{vcov.discrpar}
\title{Extract Discrimination Parameters of Item Response Models}
\description{
A class and generic function for representing and extracting the
discrimination parameters of a given item response model.
}
\usage{
discrpar(object, \dots)
\method{discrpar}{raschmodel}(object, ref = NULL, alias = TRUE, vcov = TRUE, \dots)
\method{discrpar}{rsmodel}(object, ref = NULL, alias = TRUE, vcov = TRUE, \dots)
\method{discrpar}{pcmodel}(object, ref = NULL, alias = TRUE, vcov = TRUE, \dots)
\method{discrpar}{plmodel}(object, ref = NULL, alias = TRUE, vcov = TRUE, \dots)
\method{discrpar}{gpcmodel}(object, ref = NULL, alias = TRUE, vcov = TRUE, \dots)
}
\arguments{
\item{object}{a fitted model object whose discrimination parameters should be
extracted.}
\item{ref}{a restriction to be used. Not used for models estimated via CML as
the discrimination parameters are fixed to 1 in \code{raschmodel}s,
\code{rsmodel}s and \code{pcmodel}s. For models estimated via MML
(\code{plmodel}s and \code{gpcmodel}s), the parameters are by default
identified via the distributional parameters of the person parameters (mean and
variance of a normal distribution). Nevertheless, a restriction on the ratio
scale can be applied.}
\item{alias}{logical. If \code{TRUE} (the default), the aliased parameters are
included in the return vector (and in the variance-covariance matrix if
\code{vcov} = TRUE). If \code{FALSE}, these parameters are removed. For
\code{raschmodel}s, \code{rsmodel}s and \code{pcmodel}s where all
discrimination parameters are fixed to 1, this means that an empty
numeric vector and an empty variance-covariance matrix is returned if
\code{alias} is \code{FALSE}.}
\item{vcov}{logical. If \code{TRUE} (the default), the variance-covariance
matrix of the discrimination parameters is attached as attribute
\code{vcov}.}
\item{\dots}{further arguments which are currently not used.}
}
\details{
\code{discrpar} is both, a class to represent discrimination parameters of
item response models as well as a generic function. The generic function can
be used to extract the discrimination parameters of a given item response
model.
For objects of class \code{discrpar}, several methods to standard generic
functions exist: \code{print}, \code{coef}, \code{vcov}. \code{coef} and
\code{vcov} can be used to extract the discrimination parameters and their
variance-covariance matrix without additional attributes.
}
\value{
A named vector with discrimination parameters of class \code{discrpar} and
additional attributes \code{model} (the model name), \code{ref} (the items or
parameters used as restriction/for normalization), \code{alias} (either
\code{TRUE} or a named numeric vector with the aliased parameters not included
in the return value), and \code{vcov} (the estimated and adjusted
variance-covariance matrix).
}
\seealso{\code{\link{personpar}}, \code{\link{itempar}},
\code{\link{threshpar}}, \code{\link{guesspar}}, \code{\link{upperpar}}}
\examples{
o <- options(digits = 4)
## load verbal aggression data
data("VerbalAggression", package = "psychotools")
## fit Rasch model to verbal aggression data
rmod <- raschmodel(VerbalAggression$resp2)
## extract the discrimination parameters
dp1 <- discrpar(rmod)
## extract the standard errors
sqrt(diag(vcov(dp1)))
if(requireNamespace("mirt")) {
## fit 2PL to verbal aggression data
twoplmod <- plmodel(VerbalAggression$resp2)
## extract the discrimination parameters
dp2 <- discrpar(twoplmod)
## this time with the first discrimination parameter being the reference
discrpar(twoplmod, ref = 1)
## extract the standard errors
sqrt(diag(vcov(dp2)))
}
options(digits = o$digits)
}
\keyword{classes}
|
