graphics: Graphic Functions to Illustrate Response Curves and Parameter...

In irtProb: Utilities and Probability Distributions Related to Multidimensional Person Item Response Models

Description

Graphic functions to illustrate response curves and parameter estimation.

Usage

PCC(theta  = 0, S = 0, C = 0, D = 0,
    s      = 1/1.702, b = seq(-5, 5, length = 300), c = 0,
    d = 1, groups = TRUE, ID = "ID",
    main   = "Person Characteristic Curve",
    xlab   = "Item Difficulty Parameter (b)",
    ylab = "P(x = 1)", type   = c("g", "a"))
 

Arguments

theta	numeric; vector of person proficiency (θ) levels scaled on a normal z score.
S	numeric: positive vector of personal fluctuation parameters (σ).
C	numeric: positive vector of personal pseudo-guessing parameters (χ, a probability between 0 and 1).
D	numeric: positive vector of personal inattention parameters (δ, a probability between 0 and 1).
s	numeric: vector of item fluctuation parameter or the inverse of item discrimination (s= 1/a).
b	numeric: vector of item discrimination parameter.
c	numeric: vector of item pseudo-guessing parameter.
d	numeric: vector of item inattention parameter.
ID	character: curves identification information displayed ("ID", "ALL", "THETA2 or NULL)
groups	logical: default to TRUE. If TRUE, Lattice xyplot by groups. If FALSE, xyplot with shingles.
main	character: first line of main title.
xlab	character: label of x axis.
ylab	character: label of y axis.
type	character: type of xyplot graphic. One of the following: "p", "l", "h", "b", "o", "s", "S", "r", "a", "g", "smooth".

Value

PCC returns a list:

graphic	trellis object: figures for each subject (group or shingle representation).
probability	data.frame: item snd person parameters, like th eprobability of a correct response.

Author(s)

Gilles Raiche, Universite du Quebec a Montreal (UQAM),

Departement d'education et pedagogie

[e-mail], http://www.er.uqam.ca/nobel/r17165/

Examples

## PCC curves grouped on a single figure
 res1 <- PCC(theta=c(-2,-2,-2),S=0, C=c(0.0, 0.1, 0.6), D=0.2,
             b=seq(-5,5,length=3000), ID=NULL, groups=TRUE,
             type=c("g","a"))
 res1
 
## PCC curves shingled on a single figure for each subject
 res2 <- PCC(theta=c(-2,-1,0),S=c(4.0,0.0, 1.0), C=c(0.0, 0.1, 0.6), D=0.2,
             b=seq(-5,5,length=3000), ID=NULL, groups=FALSE,
             type=c("g","a"))
 res2
 

irtProb documentation built on May 2, 2019, 1:30 p.m.