How to analyse ordinal data in an ordinal fashion
Jan Lanke
University of Lund
Abstract
Given the task of analysing a set of ordinal data, the easy way out
is to behave as if they were interval data and perform the analysis
accordingly, hoping that one will not go too far wrong. Object of my
talk will be to discuss what can be done if one wants to stick to
ordinal-data methods; in particular I shall concentrate on the case
where the data are what, for want of a better name, I call continuous
ordinal data: such ordinal data where the set of possible values is so
large that a contingency table is not a very good way of presenting
the data.