item_characteristic_curve

# ItemResponsePlots.item_characteristic_curve — Function.

item_characteristic_curve(model, item, response)
item_characteristic_curve(model, item)

Create a plot of the item characteristic curve for response of item.

An item characteristic curve plots the probability for response of item given the person ability θ.

If response is omitted, the default plot behaviour depends on model:

• For models where response_type(model) == Dichotomous the item characteristic curve for response = 1 is plotted, i.e. the probability of a correct response.

• For models where response_type(model) <: Union{Nominal, Ordinal} all category characteristic curves are plotted.

Plot attributes

Generic

• color: The color of the item characteristic curve.

• uncertainty_color: The color of the displayed uncertainty information. For plots with uncertainty intervals this is the color of the confidence band. For plots with sample based uncertainty information this is the line color of the samples.

• theta: The values of theta for which to plot the item characteristic curve. default: -3.0:0.01:3.0.

• show_data: Overlay observed data in the plot. Ability estimates are aggregated according to bins. For models with estimation_type(model) == SamplingEstimate the posterior means of ability estimates are used for binning. default: false.

• bins: The number of bins for the observed data. default: 6.

Specific

Models with SamplingEstimate

• samples: The number of samples to plot. default: 1000

• uncertainty_type: Changes how the uncertainty of the estimate is displayed. If uncertainty_type = :samples, then iterations from the MCMC estimation are plotted. If unvertainty_type = :interval, then uncertainty intervals are plotted. default: :samples

• quantiles: The lower and upper quantile for uncertainty intervals. default: (0.1, 0.9)

• aggregate_fun: A function that aggregates MCMC samples. The provided function must take a vector as input and output a scalar value. If aggregate_fun = nothing no aggregate is plotted. default: mean

source

From ItemResponseFunctions.jl

Dichotomous items

using CairoMakie
using ItemResponsePlots
using ItemResponseFunctions

item = (a = 1.87, b = 0.22)

item_characteristic_curve(TwoPL, item)

using CairoMakie
using ItemResponsePlots
using ItemResponseFunctions

item = (a = 0.65, b = -0.3)

item_characteristic_curve(TwoPL, item, 0)

Polytomous items

using CairoMakie
using ItemResponsePlots
using ItemResponseFunctions

item = (a = 0.9, b = 0.0, t = (0.0, -0.2, 1.0))

item_characteristic_curve(GPCM, item)

using CairoMakie
using ItemResponsePlots
using ItemResponseFunctions

item = (a = 1.8, b = 0.1, t = (0.0, 1.0))

item_characteristic_curve(GPCM, item, 2)