plot.irt.prob: Plot Item Stuff

Description

This function plots item response curves/surfaces using the lattice package in addition to vector plots, contour plots, and level plots for multidimensional items, and comparison plots for examining parameter drift.

Usage

## S3 method for signature 'irt.pars'
# S3 method for irt.pars
plot(x, y, ..., type, separate, combine, items, item.names, 
    item.nums, panels, drift, groups, grp.names, sep.mod, 
    drift.sd, save.hist)
## S3 method for signature 'irt.prob'
# S3 method for irt.prob
plot(x, y, ..., type, separate, combine, items, item.names, 
    item.nums, panels, save.hist)
## S3 method for signature 'sep.pars'
# S3 method for sep.pars
plot(x, y, ..., type, separate, combine, items, item.names, 
    item.nums, panels, save.hist) 
## S3 method for signature 'list'
# S3 method for irt.pars
plot(x, y, ..., type, separate, combine, items, item.names, 
    item.nums, panels, drift, groups, grp.names, sep.mod, 
    drift.sd, save.hist)

Arguments

object of class irt.prob, irt.pars, or sep.pars. For the later two classes, probabilities are computed using the function mixed before plotting the curves.

this is an argument in the generic plot function, but is not used for these methods

…

further arguments passed to or from other methods

type

type of plot to produce (for multidimensional models). Values can include "wireframe", "contourplot", "levelplot", "vectorplot1", "vectorplot2", "vectorplot3". The default if type is missing is "wireframe". The first three types of plots are based on the correspondingly named lattice plots; properties can be controlled using arguments associated with the respective methods. See the details below for more information on multidimensional plots.

separate

logical value. If TRUE, plot the item category curves or surfaces for polytomous items in separate panels

combine

vector identifying the number of response categories to plot in each panel. If NULL, the curves will be grouped by item. combine is typically used to plot curves for more than one item in a panel.

items

numeric vector identifying the items to plot

item.names

vector of item names for use in labeling the panels

item.nums

logical value. If TRUE, include item numbers on the vector plots.

panels

number of panels to display in the output window. If the number of items is greater than panels, the plots will be created on multiple pages. If NULL, all panels will be plotted on a single page.

drift

character vector identifying the plots to create to examine item parameter drift. Acceptable values are "a", "b", "c" for the various parameters respectively, "pars" to compare all of these parameters, "TCC" to compare test characteristic curves, "ICC" to compare item characteristic curves, or "all" to produce all of these plots.

groups

numeric vector identifying the groups in x that should be included for comparing parameter drift. The values should correspond to the group number of the lowest group of each pair of adjacent groups in x.

grp.names

character vector of group names to use when creating the drift plots

sep.mod

logical value. If TRUE use different markers in the drift plots to identify parameters related to different item response models

drift.sd

numeric value identifying the number of standard deviations to use when creating the perpendicular confidence region for the drift comparison plots. If missing, drift.sd will default to 3.

save.hist

logical value. If FALSE, the plot history should be deleted (only applies to Windows); otherwise, any previously saved plots will be retained.

Details

All of the plots, with the exception of the vector plots, are based on the lattice package. For the unidimensional plots, any arguments associated with the xyplot function can be included. For the multidimensional plots, any arguments associated with the wireframe, contourplot, or levelplot functions (corresponding to the argument specified in type) can be passed as well. For the multidimensional plots, vector plots are limited to two dimensions. The other three types of plots can be used with multiple dimensions by plotting the results for the first two dimensions conditional on each combination of theta values for the other dimensions. The function can handle up to ten dimensions; however, the usefulness of the plots will likely diminish if more than four dimensions are modeled. There are three types of vector plots. The first two types create a set of vectors that originate at the location of the multidimensional difficulty and characterize the length of the vectors at the appropriate angles, relative to the axes, based on the multidimensional discrimination. Setting type equal to "vectorplot2" will also include the reference composite based on an eigenvalue decomposition of the slope parameters. The third type of vector plot is similar to the first with the key distinction that the lines are drawn from the origin to the points of multidimensional difficulty. When there are more than two dimensions, a panel of vectorplots is created for each pair of dimensions.

References

Weeks, J. P. (2010) plink: An R package for linking mixed-format tests using IRT-based methods. Journal of Statistical Software, 35(12), 1--33. URL http://www.jstatsoft.org/v35/i12/

Examples

Run this code

###### Unidimensional Examples ######
# Compute probabilites for three dichotomous (3PL) items and two polytomous
# (gpcm) items then plot the item characteristic/category curves
dichot <- matrix(c(1.2, .8, .9, 2.3, -1.1, -.2, .24, .19, .13),3,3)
poly <- matrix(c(.64, -1.8, -.73, .45, NA, .88, .06, 1.4, 1.9, 2.6),
  2,5,byrow=TRUE)
pars <- rbind(cbind(dichot,matrix(NA,3,2)),poly)
cat <- c(2,2,2,4,5)
pm <- as.poly.mod(5, c("drm","gpcm"), list(1:3,4:5))
x <- mixed(pars, cat, pm) 
plot(x)

# Compute probabilites for six items using the 3PL and plot all of the item
# characteristic curves on a single panel using the combine argument
a <- c(.71,.96,.36,1.05,1.76,.64)
b <- c(-.16,1.18,2.03,1.09,.82,-1.56)
c <- c(.22,.17,.24,.08,.20,.13)
theta <- seq(-3,3,.2)
pars <- cbind(a,b,c)
x <- drm(pars,theta)
plot(x,combine=6,item.names="Items 1-6",auto.key=list(space="right"))

# Compute probabilties for the nominal response model using the ACT
# mathematics data. Plot the item category curves for a subset of 9 items.
x <- nrm(act.nrm[[1]], rep(5,60))
plot(x, items=c(2,7,12,20,35,41,48,57,60),auto.key=list(space="right"))

# Create irt.pars object with two groups (all dichotomous items)
# rescale the item parameters using the Stocking-Lord linking constants
# Create drift plots 
pm <- as.poly.mod(36)
x <- as.irt.pars(KB04$pars, KB04$common, cat=list(rep(2,36),rep(2,36)), 
  poly.mod=list(pm,pm))
out <- plink(x, rescale="SL", base.grp=2, D=1.7)
plot(out$pars, drift="all", grp.names=c("Form X","Form Y"), 
item.nums=TRUE)

## Not run: ------------------------------------
# # Compute linking constants for six groups using mixed-format items 
# # Create drift plots with distinct markers for the two response models
# pm1 <- as.poly.mod(41, c("drm", "gpcm"), reading$items[[1]])
# pm2 <- as.poly.mod(70, c("drm", "gpcm"), reading$items[[2]])
# pm3 <- as.poly.mod(70, c("drm", "gpcm"), reading$items[[3]])
# pm4 <- as.poly.mod(70, c("drm", "gpcm"), reading$items[[4]])
# pm5 <- as.poly.mod(72, c("drm", "gpcm"), reading$items[[5]])
# pm6 <- as.poly.mod(71, c("drm", "gpcm"), reading$items[[6]])
# pm <- list(pm1, pm2, pm3, pm4, pm5, pm6)
# x <- as.irt.pars(reading$pars, reading$common, reading$cat, pm, base.grp=4)
# out <- plink(x, rescale="HB")
# plot(out$pars, drift=c("a","b"), sep.mod=TRUE)
## ---------------------------------------------

###### Multidimensional Examples ######
# Compute response probabilities for 20 items from the 
# reckase9 data. Probabilites are modeled using the M2PL. 
# Each of the six plot types will be created.
x <- drm(reckase9[[1]][[1]][31:50,],dimensions=2)

# Wireframe plot
plot(x, drape=TRUE, item.names=paste("Item",31:50))
# Contour Plot
plot(x, type="contourplot",labels=FALSE,cuts=20)
# Level Plot
plot(x, type="levelplot",cuts=20)

# Use all the items for the vector plots
x <- drm(reckase9[[1]][[1]],dimensions=2)
# Vector Plot 1
plot(x, type="vectorplot1",item.nums=FALSE)
# Vector Plot 2
plot(x, type="vectorplot2",item.nums=FALSE)
# Vector Plot 3
plot(x, type="vectorplot3",xlim=c(-1.5,1.5),ylim=c(-1.5,1.5))


# Compute response probabilities for a single three-category item using 
# the multidimensional generalized partial credit model for three 
# dimensions. Plot the conditional item category surfaces in a single 
# panel and then the separated item category surfaces in separate panels
pars <- matrix(c(1.1999,0.5997,0.8087,2.1730,-1.4576),1,5)
x <- gpcm(pars,3,dimensions=3,theta=-4:4)

# plot combined item category surfaces. Rotate the plots using the
# screen argument
plot(x, screen=list(z=-30,x=-60))

## Not run: ------------------------------------
# # plot separated item category surfaces
# x <- gpcm(pars,3,dimensions=3)
# plot(x,separate=TRUE,drape=TRUE,panels=1)
## ---------------------------------------------