GitHub

RIrtWrappers.jl

This package wraps some R packages for fitting IRT models and running CATs.

Fitting IRT models

The KernSmoothIRT and Mirt modules wrap R libraries for fitting IRT models.

The models are returned as item banks as in FitttedItemBanks.jl. The inputs are response matrices, in the same format as provided by ItemResponseDatasets.jl. In particular a DataFrames.DataFrame with questions as columns and respondents as rows, with outcomes 0-based integer coded.

For example:

4×4 DataFrame
   Row │ Q1     Q2     Q3     Q4
       │ Int64  Int64  Int64  Int64
───────┼───────────────────────────
     1 │     1      1      0      0
     2 │     1      0      1      0
     3 │     1      1      1      1
     4 │     1      1      0      1

Running CATs

The MirtCAT module wraps an R library for simulating and administering CATs.

You need to use the following method to get ahold of the module.

RIrtWrappers.require_mirtcat — Function
require_mirtcat() -> Module

Returns the MirtCAT extension module. Requires the ComputerAdaptiveTesting module in your environment.

source

The main objects from these always implement the ComputerAdaptiveTesting.Stateful interface.

Index