icc: Item Characteristic Curve

In duckmayr/bggum: Bayesian Estimation of Generalized Graded Unfolding Model Parameters

Description

Plots item characteristic curves given alpha, delta, and tau parameters.

Usage

icc(a, d, t, from = -3, to = 3, by = 0.01, layout_matrix = 1,
  main_title = "Item Characteristic Curve", sub = "",
  color = "black", plot_responses = FALSE, thetas = NULL,
  responses = NULL, response_color = "#0000005f")

Arguments

a	A numeric vector of alpha parameters
d	A numeric vector of delta parameters
t	Either a list of numeric vectors for the tau parameters for each option, or a numeric vector if the IRF for only one item is desired – note the first element of each vector should be zero
from	A numeric vector of length one, the lowest theta value to estimate response probabilities for; default is -3
to	A numeric vector of length one, the highest theta value to estimate response probabilities for; default is 3
by	A numeric vector of length one giving the spacing between theta values; default is 0.01
layout_matrix	An integer matrix dictating the layout of the plot; the default is a one-column matrix with one element for each item
main_title	A character vector giving the plots' main titles; default is "Item Characteristic Curve".
sub	An optional character vector of subtitles for the resulting plots, to be pasted onto the main title (helpful for titling individual plots when plotting multiple items' ICCs).
color	The color to plot the ICC line in; default is "black"
plot_responses	A logical vector of length one specifying whether to draw points at the theta estimates of actual responses; default is FALSE
thetas	An optional vector of theta estimates for response drawing; if plot_responses = TRUE and thetas is not provided, an error will be thrown.
responses	An optional matrix or vector (if the ICC for only one item is desired) of responses; if plot_responses = TRUE and responses is not provided, an error will be thrown. NOTE: The lowest response for each item should be 0, not 1.
response_color	The color to plot the response points when plot_responses = TRUE; the default is "#0000005f".

Examples

## We'll simulate data to use for these examples:
set.seed(123)
sim_data <- ggum_simulation(100, 10, 4)
## You can plot the ICC for one item:
icc(sim_data$alpha[1], sim_data$delta[1], sim_data$tau[[1]])
## Or multiple items:
icc(sim_data$alpha[1:2], sim_data$delta[1:2], sim_data$tau[1:2], sub = 1:2)
## You can also plot the actual responses over the expected response line:
icc(sim_data$alpha[1], sim_data$delta[1], sim_data$tau[[1]],
    plot_responses = TRUE, responses = sim_data$response_matrix[ , 1],
    thetas = sim_data$theta)

duckmayr/bggum documentation built on Jan. 20, 2020, 5:23 a.m.