ABSTRACT
Up to now we have dealt exclusively with two-way tables in which the frequencies of co-occurrence of two variables have been mapped. We now start to consider situations where data are available on more than two variables and how we can explore such data graphically. One approach is to re-format the multiway frequency table as a two-way table and to use the usual simple correspondence analysis (CA) approach. The various ways of reformatting the table give different viewpoints of the association patterns in the data.
