(Musenm of Natural History, University of Oregon, 1968-09) Johnson, LeRoy Jr.
There are two basic needs
which this paper may help to meet. First, seriation
is a current issue, or at least a current
method, particularly in the literature of anthropology
and archaeology. The topic has regrettably
never been given ample discussion and
the need for such is self-evident. Second, seriation
is a logical point at which students of biology,
natural history, and social science may be
introduced to the useful techniques of scale
analysis and cluster analysis, to build an understanding
of several of the problems involved
in such studies before confronting more elegant
techniques of analysis. It is hoped that the sophisticate
in numerical analysis will not be impatient,
then, with the following exposition,
since it is addressed to research workers and
students having a minimum acquaintance with
numerical analysis.
The following discussion will treat seriation
from several standpoints, notably (1) its basic
nature, (2) its history and treatment in the
literature, (3) its problems for computer processing,
and (4) its proper use, particularly
in conjunction with other aids, in generating
overviews for bodies of numerical data.
