Firestar-D: Computerized Adaptive Testing Simulation Program for Dichotomous Item Response Theory Models

摘要

Computerized adaptive testing (CAT) enables efficient and flexible measurement of latent constructs. The majority of educational and cognitive measurement constructs are based on dichotomous item response theory (IRT) models. An integral part of developing various components of a CAT system is conducting simulations using both known and empirical data. Having access to open-source CAT algorithms, which do not utilize proprietary optimization routines, can be pedagogically instructive and can facilitate the understanding and advancement of the science behind CAT. Firestar-D is a computer program for simulating CAT for dichotomous IRT models under the three-parameter IRT model framework (Birnbaum, 1968). The program supports a range of alternatives for item selection (Choi & Swartz, 2009; van der Linden, 1998), content balancing (Kingsbury & Zara, 1989), exposure control (Revuelta & Ponsoda, 1998), provisional and final theta estimation, stopping, and output control. Firestar-D was developed using the same platform as its antecedent, Firestar for polytomous models (Choi, 2009): The back-end engine of the program was developed in R (http://cran.r-project.org/) and the front-end user interface in C#. R is a system for statistical computation and graphics and is freely available under the GNU (GNU General Public License v3.0: http://www.gnu.org/licenses/gpl.html) "copy-left" license agreement. The Firestar-D program is designed to run on Windows-based computers with R installed.