11.1 Simulation study in CDM
In Cognitive Diagnostic Modeling (CDM), a simulation study refers to the use of computer-generated data to evaluate the performance of cognitive diagnostic models or methods.
CDMs are used to classify individuals based on their mastery of specific skills or attributes, and a simulation study in this context allows researchers to understand how well a CDM performs under various conditions.
Simulation studies are a core part of research in the psychometric literature.
A simulation may also be referred to as a computer simulation or a Monte Carlo (MC) simulation.
The popularity and utility of simulations suggests that researchers and practitioners in measurement should be familiar with their appropriate use, the kinds of questions that simulations can answer best, and the most effective way to present or communicate simulation results(Feinberg & Rubright, 2016)
One advantage of simulation studies is that they allow researchers to compare estimated parameters against their respective true parameters, which are unknown for real data applications
Why do we need simulation study in CDM?
Model Evaluation
Parameter Recovery
Performance under Different Conditions
Comparison of Models
Testing Model Assumptions