1. Why do we need to estimate the accuracy for the empirical classification? Empirical classification is an estimate as well.
2. Classification accuracy is a kind of estimate for a population of examinees who can pass or fail a test. Is it more important than when mistakenly classifying an examinee?
3. In simulation study, the estimation of CA was based on true item parameter and distribution of theta. It may be more critical to check the robustness of Lee and Rudner approach.