The problem of misuse of statistical packages and hence statistical methods
has long peen recognised but no real solution has yet appeared, although some
useful work has been done in related fields. This problem appears to lend itself
to expert system technology.
This paper outlines the history of research in statistical expert systems, showing
how the demands of data analysis are different from those of other fields.
Furthermore, the paper reviews some recent work and discusses the development
of an expert system for the planning of experiments.