Questions and Comments:
1. In mixture models, the label switch is common. It was found that the reversed KL information was found to function similarly in the study. Is it the reason that the reversed KL information is directly related to the label switch?
2. In this study, the criterion of maximizing the distinction between latent classes is used to select the next item in Method 1, 3 and 4. Under this criterion, the items with larger difficulty differences are selected to be administered. However, when the content balance is taken into account, the criterion is challenging.
3. The overall classification accuracy was reported in the present study. However, it would be more interesting to provide the cross-tab classification accuracy. That is, the percentage of correctly classifying the class 1 examinees to class 1 and, the percentage of incorrectly classifying the class 1 examinees to class 2. The same for class 2.