Chemometrics

This book is one of a series of easy-to-read, user-friendly texts written by some of the foremost lecturers in analytical chemistry in the UK. The texts are designed for training, continuing education and updating of all technical staff concerned with analytical chemistry. It is aimed at those interested in analytical chemistry and instrumental techniques who wish to study in a more flexible way than traditional institute attendance or to augment such attendance. This text concerns experimental design. Prat 1 provides a revision of some of the statistical methods used in experimental design and part 2 introduces randomization, replication and blocking, followed by some designs which have increasing degrees of complexity as far as blocking is concerned. These are some of the more fundamental areas of experimental design. Part 3 introduces designs with more than one factor (factorial designs) and mainly covers the two-level designs, which are useful for discovering whether variables are important. The emphasis here lies on discovering whether main effects and interactions are important.
Part 4 briefly covers fractional factorial designs which are very useful exploratory designs when it is not certain whether factors have any effect upon an experimental response. These designs have the advantage over complete factorial designs when a large number of factors has to be included in an initial investigation since they use a relatively small number of experimental runs. Part 5 covers response surface methodology.