Statistical aspects of chemical analysis

Following a brief statement of certain fundamental principles, methods of sampling are dicussed which enable the informaiton sought by the analyst to be obtained as economically as possible. The need for proper randomisation both in sampling and in performing analyses is stressed and a short account is given of the use of 'quality control' charts in analytical work. Attention is drawn to the possibility of estimating the standard error from the range. A method of extracting the maximum information from analytical investigation involving the estimation of a ratio is described. The advantages of the 'factorial' as opposed to the 'classical' type of experimental design are demonstrated and reference is made to the value of carrying out an analysis of variance when examining the experimental data.