ABSTRACT
Most of the currently used item response models are designed for testing situations, where test takers are presented with multiple, written items that are usually (not necessarily) scored dichotomously. A particularly well-known model for these kinds of test data is the one-parameter logistic (1-PL) or Rasch model (Volume One, Chapter 3). In such a model, the chance of a test taker answering an item correctly is a function of a person ability parameter and an item difficulty parameter as in 1-PL. Not all tests follow this item format and some of them require another type of modeling.
