A supervised learning approach to estimating IRT models in small samples
{{output}}
Existing estimators of parameters of item response theory (IRT) models exploit the likelihood function. In small samples, however, the IRT likelihood oftentimes contains little informative value, potentially resulting in biased and/or unstable parameter estima... ...
