Chapter 10 - Applied Statistics in Product Development

Abstract

There are only three kinds of mathematical data: nominal, ordinal, and numerical. For each type of data different statistical techniques and methods of evaluation have been devised. Nominal data are obtained by conducting a census or survey. Ordinal data are actually rarely obtained in practice in pharmaceutical process research and development. To truly collect ordinal data, one has to compare each and every member of the group of objects being ranked with every other member of the group. Numerical data are generally used by scientists and engineers. A device is used with a measurement scale, typically graduated in equal intervals, to obtained so-called interval scale numerical data. Objects measured on such a scale are larger or smaller than other objects by some additive factor. Such data can include both positive and negative real numbers, since the zero point of an interval scale can be arbitrarily chosen. Other numerical data are measured on a so-called ratio scale. Objects measured on such a scale are proportionally larger or smaller than other objects by some multiplicative factor. Such data can include only positive real numbers, since the zero point of a ratio scale is fixed absolutely.