ABSTRACT
Item response theory is a powerful scaling technique that uses a mathematical model to depict the probability of a correct response given item and person parameters. Model fit at the item level is often assessed by comparing the observed proportions of examinees responding to a particular category to predictions based on the model being applied. The set of observed proportions is used as a proxy for the response function that is not constrained to follow any particular shape. Graphical techniques are useful for portraying evidence of model fit in a convenient and informative display. Although residuals provide useful information for evaluating item-level model fit, they are difficult to interpret in tabular form, especially when analyzing many items. Model fit plays an important role in the valid application of Item Response Theory models.
