The application of a class of Rasch models is discussed for situations in which test items are grouped into subsets and the common attributes of items within the subsets bring into question the usual assumption of conditional independence. Models are expressed as particular cases of the random coefficients multinomial logit model. (SLD)

Marginal maximum likelihood estimation equations are derived for the structural parameters of the Saltus model, and a computing approximation is suggested based on the EM algorithm. The solution is illustrated with simulated data and an example from the domain of mixed number subtraction. (SLD)

A strategy is described for dealing with measurement situations in which certain categories of responses are null, that is, persons do not respond in certain categories to certain items. The method is described for the partial credit model while maintaining the integrity of the original response framework. (SLD)