Evaluate model performance

Measure model performance through fit, comparisons, and reliability.

Introduction

The second article will focus on tools for evaluating an estimated DCM model. This will include tools for calculating:

• Absolute fit (e.g., fit_ppmc(), fit_m2())
• Relative fit for model comparisons (e.g., waic(), loo())

We will showcase tools for investigating potential causes of misfit such as:

• Incorrect Q-matrix specifications (qmatrix_validation())
• Local item dependence (yens_q3())
• Item and attribute discrimination (cdi())