Statisticians have developed many graphical displays for continuous data. By contrast, displays for categorical data have been rare. An exception are "mosaic plots," proposed in 1981 by Hartigan and Kleiner. Although mosaic plots form a powerful recursive visualization technique, they have not proven popular. The reason for their lack of popularity is the fact that their visual impact depends on an ordering of the variables. This disadvantage is critical in static implementations of mosaic plots, such as in SAS or S-Plus. In interactive implementations, however, software can offer very flexible means of rearranging the order of the variables, both manually and automatically. This has been implemented in the MANET system for interactive data visualization. In addition, mosaic plots and barcharts lend themselves to interactive linked highlighting in multiple views. I show how this facilitates visualization of categorical models. This can be achieved by superimposing residual information for loglinear models, as well as by simple linked highlighting for response models. Users can thus perform graphical stepwise modeling of categorical data. This represents a new methodology that goes beyond traditional exploratory uses of data visualization. In this talk I give an overview of various visualization techniques for categorical data and demonstrate the incorporation of mosaic plots in MANET. For more information on MANET, point your browser to MANET.