Sampling techniques and their application to managerial problems

In future managers will depend more on numerical information in making decisions, and this information will be more readily available with the extension of the use of computers. Some of the decisions made by management are of necessity based on information which, although expressed numerical l7 is usually doubtful to a greater or lesser extent. The theory of probability provides a means of quantifying the degree of doubt in mathematical terms. The basic concepts of sampling techniques are described and although a manager does not necessarily have to be thoroughly acquainted with the mathematical and statistical content, he should at least know what techniques are available and how they should be used. The measurements of central tendency and dispersion are described, together with the Normal and Negative exponential distributions. The measurements of the standard error of the means for both small and large samples are given and exemplified. A method of determining the number of observations required in a Time Study is introduced. The general theory of significance tests is described and illustrated with examples. t and x2 tests are described and the corresponding tables are appended. The use of single and bivariate proportionate sampling techniques is illustrated, and the modifications required when the costs of the respective strata samples are incorporated. The significance of symmetrical and asymmetrical confidence regions is shown. Examples are given of the application of statistical quality control and also of Sequential Analysis.