Procedures Based on Item Response Theory Models for the Development of Short Test Forms

Implement different Item Response Theory (IRT) based procedures for the development of tests from item bank. The procedures are flexible enough to be adopted for the development of short forms of full-length tests. Different procedures are considered (Epifania, Anselmi & Robusto, 2022 and Epifania & Finos, 2025 ). The main difference between the presented procedures refers to the degree of control that they allow for targeting specific latent trait levels. The simplest procedure, denoted as benchmark procedure, does not allow for any control on the latent trait levels of interest, while the other procedures allow for specifying either discrete latent trait levels for which the information needs to be maximized (theta-target procedure, ) or a target information function that needs to be recreated with the selected items (item selection algorithm -ISA- denoted as Frank in ). Another difference concerns the definition of the number of items to be selected. In the benchmark and theta-target procedures, the number of items must be defined a priori, while in ISA the number of items is determined automatically by the algorithm.

shortIRT

The goal of shortIRT is to simple tool for the development of static short test forms (STFs) in an Item Response Theory (IRT) based framework. Specifically, two main procedures are considered:

1. The typical IRT-based procedure for the development of STFs (here denoted as benchmark procedure, BP), according to which the most informative items are selected without considering any specific level of the latent trait

2. The IRT procedure based on the definition of levels of interest of the latent trait (i.e., $\theta$ targets, here denoted as $\theta$-target procedure). The selected items are those most informative in respect to the $\theta$ targets. This procedure can be further categorized according to the methodology used for the definition of the $\theta$ targets:

   1. Equal interval procedure (EIP): The latent trait is divided in $n + 1$ (where $n$ is the number of items to be included in the STF) intervals of equal width and the central points of each interval are the $\theta$ targets
   2. Unequal interval procedure (UIP): The latent trait is clustered in $n$ clusters (where $n$ is the number of items to be included in the STF) and centroids of each clusters are the $\theta$ targets
   3. User defined procedure (UDP): The user manually defines the $\theta$ targets to which the STF should tend to. They might also be the same $\theta$ values (e.g., for the development of a screening STF with a cut-off point).

Installation

You can install the development version of shortIRT from GitHub with:

# install.packages("devtools")
devtools::install_github("OttaviaE/shortIRT")