I just wanted to know if it was possible to get IRT scores estimates for a single subject starting from a previously estimated IRT model.
Imagine that an IRT model is fitted on a standardization sample in order to estimate the items parameters. Then someone answer the previously calibrated test. Can we use the responses pattern of this person and estimate a single individual score through the known item parameters ? Should we add the data of this individual to the sample and re-estimate the model (but the item parameters may slightly change) or is there an easiest way to do that ?