Although the study of human values is excluded by classical science, i
t is a major area of interest in systems science. Its difficulties cha
llenge much of the field's creative force. This manifests itself in a
strong tendency to propose what is 'new': directions of research, meth
odologies, paradigms. This variety poses problems of evaluation. A too
l is described to solve these problems. It simplifies the process of c
omparing studies, and suggests how to improve on them.