Hi Nikolay,
Thanks again for the additional information. I thought there may have been more substantial differences between the marginal and joint MLE algorithms for estimating the item and person parameters, but this definitely helps quite a bit.
Thanks again,
Billy