The advent of the laboratory computer
has allowed analytical chemists to col-
lect enormous amounts of data on a
wide variety of problems of interest.
With this ability, however, has come
the realization that more data does not
necessarily mean more information.
One can collect reams of computer out-
put without knowing anything more
about the system under investigation.
Only when theda t a a r e int e rpr e t edand
put to use do they become valuable to
the chemist and t o society in general:
thendata become information. For this
reason, data analysis methods that can
handle large quantities of data are be-
comine a necessitv for the nracticine
the awareness of the need for moreso-
phisticated techniques grows. Howev-
er, widespread use of the full set of ana-
lytical tools will be realized only when
the analyst is familiar with the general
goals and advantages of the methods.
a n y situation where multiple measure-
menta are acquired.
One example of a n analytical prob-
lem solved by near-IR reflectance anal-
ysis is the estimation of the protein and
moisture content of wheat samnles. As
~
introduction to the area of multivariate anal-
yses known as multivariate calibration.
Our goal is to give the chemist insight
into the workings of a collection of statistical techniques and to enable him or
her to judge the appropriateness of the
techniques for an application of interest. The list of methods described is by
no means a complete representation of
those tha t can he applied to multivari
ste data. Also, we do not compare the
results obtained using one method over
another; instead, we describe the methods in the line of a tutorial. For this
reason, the examples chosen illustrate
how the methods work rather than
with many analytical methois, this
procedure consists of two phases: cali
bration and prediction. T h e chemist
begins by constructing a data matrix
(R) from the instrument responses a t
selected wavelengths for a given set of
calibration samples. I n the case of
near-IR reflectance analysis, the R ma-
trix can be constructed from the loga-
rithmic reflectances (log(l/ref)) or
from some other combination of reflec-
tances obtained a t the various wave-
lengths ( 4 ) . A matrix of concentration
values (C) is then formed using inde-
pendent or referee methods such as
Kjeldahl nitrogen analysis for protein
determination and oven-drvine
. - for the
compareamony them. The threemulti-
variate methods that will be discussed
determination of moisture content.
The goal of the calibration Dhase is to
are multiple linear regression (MLR),
principal component regression (PCR),
and partial least squares (PLS). MLR,
a n extension of ordinary least
squares, is the easiest to understand
and also the most commonly used;
PCR and PLS have yet t o achieve
widespread acceptance in chemistry.