ABSTRACT
In the past 40 years the information made available by
chemical analysis has increased greatly due both to the
increased number of analyzed samples and the number
of analytes measured in each sample. For this last
reason chemical information is now almost always
multivariate information, where each object (sample,
molecule, panelist, individual) is described by many,
sometimes thousands, variables. At the same time, the availability of desktop
computers with continuously increasing speed and
low cost has permitted the handling of chemical
information using the tools of multivariate statistics
and applied mathematics. Two disciplines, Experimental Design: How do we
perform experiments to obtain the maximum relevant
information at the minimum cost? and Chemometrics:
How do we get chemically relevant information out of
measured chemical data, how do we represent and
display this information, and how do we get such
information into data? (1) began to be a part of the
background of many chemists. The special needs of chemistry, in particular food
chemistry, and the continuous exchange of ideas and
techniques among the two disciplines had a strong
influence on their evolution. As a consequence, we now
prefer to redefine chemometrics to include experi-
mental design, for which, according to Roger Phan-
Tan-Luu, we should use the name ‘‘methodology of
experimental research’’ to stress its importance in the
science. So chemometrics can be defined as the chemical
discipline with the following objectives:
1. To plan the chemical experiments in order to
obtain the maximum useful information with
the minimum cost. 2. To optimize the analytical process, with refer-
ence to the conditions of maximum repeatabil-
ity, reproducibility, robustness, and especially
to the step in which the chemical quantity
is computed from the measured physical quan-
tities. 3. To extract useful information, relevant to the
chemical problem, from the computed chemical
quantities and from the other available data.
