It is not an easy task to gather appropriate data and convert them to high-quality numerical information at low cost. Although they appear as separate disciplines and are taught separately at universities, accounting, computer science, operations research and statistics essentially are concerned with the solution of this problem. Separate disciplines exist only because they differ in approach or in the type of information they seek to obtain. Sometimes a new discipline is formed to study only a limited part of the management information system.

As a collection of tools to produce high-quality information economically, statistics has great potential in business and government. In statistical methods, economy and time savings are achieved by approximation. A common example is the application of statistical sampling to estimate characteristics of the environment of an organization. Sampling also is used in auditing to estimate financial information.

Recently, applications of statistical methods to model the behavior of variables have become popular as well. For example, a simple model fitted to the data of past sales might be used to generate forecasts of future sales. Statistical methods also can be used to approximate the complex relationship between a set of goal variables and some other set of policy instruments; the statistically-formulated model then is used as a guide to control future values of goal variables.

The success of a statistical method for acquiring information, be it values of variables or structures, depends upon whether or not the approximation is good enough for management decision making. That is, success depends upon the statistics' ability to remove uncertainty economically, as required by management. This cannot be evaluated unless the statistical method actually is put to work.

                                    H. Tamura
                                    University of Washington

                                    JOURNAL OF CONTEMPORARY BUSINESS VOL 8, NO.3