An empirical comparison of two cognitive diagnosis models for polytomous attributes

Paper Session, M7: Model fit issues with Diagnotic Classification ModelsThis paper proposes the fully additive model (fA-M), and compares it with the polytomous generalized deterministic input, noisy 'and' gate (pG-DINA model) using real data. Different statistics indicate that the fA-M provides a better model-data fit than the pG-DINA model . Moreover, the two models produces very disparate examinee classifications