Categorical data appear in all areas of data analysis, from social sciences
and surveys to data mining. They occur either in the form of nominal or
ordered variables and interval grouped data as in (possibly censored) data
of statistical offices. As computers and methods are able to handle ever
larger data sets, the importance of analysing categorical data grows accordingly.
Approaches are made in this direction, but often enough the analysis remains
on the level of merely a listing of numbers. Data mining plays an especially
large role, since in this field categorical data are not only analysed
but also vast amounts of categorical output are produced and have, again,
to be analysed in order to obtain interpretable results. In the field of
statistical modelling there are several approaches in dealing with multivariate
categorical data - linear and log-linear models, logit and probit models
are some of the most common methods. For all of these methods it is necessary
to check how well the data are fitted. Examining residuals with respect
to structural behaviour or irregularities is vital. In the case of continuous
data, graphical displays are used for this task. For categorical data graphical
displays, also, exist, even for high-dimensional situations. But the connection
between the graphical display and the model is far less explored for categorical
data than for continuous data.
