nd imperfect reckoning with the facts.

The person who uses scientific method takes account of all his known
forces; he prepares his materials, controls his processes and isolates
his factors so as to reveal the bearing of every step in the process
upon an ultimate and often a far distant result. In other words, he
tries at every stage to build upon a sure foundation. His trained
imagination and judgment working on known facts set the limit on what
he may expect to find, and interpret what he does find, all along the
way.

In so far as particular business enterprises have rested on
engineering, chemistry, biology, and other sciences, a scientific
method of approach has long had large use in business; but the
scientist in business has usually been a salaried expert--a man apart
from the management--and it has been his results, and not necessarily
his methods, that have influenced business practice. We are now coming
to understand that scientific method is the only sure approach to all
problems; it is a thing of universal application, and far from being
confined to the technical departments of business, where the technical
scientists hold sway in their particular specialties, it may have its
widest application in working out the problems of management.

The way in which a man trained in scientific method may determine
business practice in a scientific manner finds illustration in a
multitude of practical business problems, ranging all the way from the
simplest office detail to the most far-reaching questions of policy. To
cite an example, of the simpler sort: if an item in an order sheet is
identical for eight out of ten orders is it better to have a clerk
typewrite the eight repetitions along with the two deviations or to use
a rubber stamp? Of course, there are not one or two, but many, items in
an order sheet and the repetitions and deviations are not the same for
all items. In practical application, the rubber-stamp method means a
rack of rubber stamps placed in the most advantageous position. It
requires also a decision as to the precise percentage of repetitions
which makes the stamp advantageous. Then arises the further question,
why not have the most numerous repetitions numbered and keyed and thus
avoid the necessity of transcribing them at all?

The rule-of-thumb approach to this kind of problem would proceed from
speculations concerning the effect of interrupting the process to use
the stamp, the result of