Page 18 - Statistics and Data Analysis in Geology

Introduction

My contention is that a quantitative approach to geology can yield a fruitful re-
turn to the investigator; not so much, perhaps, by “proving” a geological hypothesis
or demonstrating its validity, but by gaining insights from the critical examination
of phenomena that is prerequisite to any quantitative procedure. Numerical analy-
sis requires that collection of data be carefully controlled, with consideration given
to extraneous influences. As a consequence, the investigator may acquire a closer
familiarity with the objects of study than could otherwise be attained. Certainly
a paleontologist who has made careful measurements on a large collection of ran-
domly selected fossil specimens has a far greater and more accurate understand-
ing of the natural variation of these organisms than does the paleontologist who
relies on informal examination. The rigor and objectivity required by quantitative
methodologies can compensate in part for insight and experience which otherwise
must be gained by many years of work. At the same time, the discipline neces-
sary to perform quantitative research will hasten the growth and maturity of the
scientist.
The measurement and analysis of data may lead to interpretations that are
not obvious or apparent when other means of investigation are used. Multivariate
methods, for example, may reveal clusterings of objects that are at variance with
accepted classifications, or may show relationships between variables where none
were expected. These findings require explanation. Sometimes a plausible explana-
tion cannot be found; but in other instances, new theories may be suggested which
would otherwise have been overlooked.
Perhaps the greatest worth of quantitative methodologies lies not in their ca-
pability to demonstrate what is true, but rather in their ability to expose what is
false. Quantitative techniques can reveal the insufficiency of data, the tenuousness
of assumptions, the paucity of information contained in most geologic studies.
Unfortunately, upon careful and dispassionate analysis, many geological interpre-
tations deteriorate into a collection of guesses and hunches based on very little
data, of which most are of a contradictory or inconclusive nature.
If geology were an experimental science like chemistry or physics-in which
observations can be verified by any competent worker-controversy and conflict
might disappear. However, geologists are practitioners of an observational sci-
ence, and the rigorous application of quantitative methods often reveals us for the
imperfect observers that we are. Indeed, a decline into scientific skepticism is one
of the dangers that often traps geomathematicians. These workers are often char-
acterized by a suspicious and iconoclastic attitude toward geological platitudes.
Sadly it must be confessed that such cynicism is often justified. Geologists are
trained to see patterns and structure in nature. Geomathematical methods provide
the objectivity necessary to avoid creating these patterns when they may exist only
in the scientist’s desire for order.

5