Graphical Tools for Item Response Theory Model Assessment

This is an implement of graphical checking tools for IRT model discussed in <Graphical Tools for Item Response Theory Model Assessment>.

1. R package:

rstan, KernSmoothIRT

2. IRT model assessment:

• simulate data from 2PL model or find a real data set
• fit data into three Item Response Theory models: Rasch, 2PL, 3PL
• use graphical tools to assess the performances of the models. The graphical tools include:
  • residual diagnostic plots based on Randomized Quantile Residual (RQR)
  • item characteristic curves (ICC) of a non-parametric model (Kernel Smoothing Model) and the posited parametric models

3. Result

The images are presented in the folder result.