Toward a Functional Accounting System ; Accounting Variances and Statistical Variance Analysis toward a Functional Accounting System; Accounting Variances and Statistical Variance Analysis

The functions of accounting and statistics are in most respects similar, These fields are both tools, concerned with the processes of measurement, collection of data, analysis and decision making. In performing their functions both accounting and statistics rely heavily on signals generated by variations from expected performance. Even in their manifested differences these two fields are at worst complementary. For example, accounting is concerned with the design of the total system which will provide the relevant information on a continuous basis, while statistics focuses more on ad hoc micro analyses such as design and analysis of experiments for eliminating hypotheses, which are aimed at determining "relevance". In view of all the similarities and complementarities, one wonders why statistical analyses have not become an integral part of the accounting systems. The purpose of this paper is to point out some ways by means of which statistical techniques can be introduced into the accounting system to author wishes to thank Kenan Sahin a research assistant at MIT for his comments. make its output more meaningful and provide flexibility for measuring the potential outcome of alternative courses of action. To do this, I will: (a) Provide some background material on my previous efforts in this direction and also elaborate a little more on what motivated me in the first place, (b) show that the output of standard accounting systems can be used for statistical analyses of variance, and (c) carry out some of the consequences of statistical analyses and of testing hypotheses for what I call functional managerial accounting systems. In an appendix, I will apply such statistical techniques to determine the statistical significance of variations in the performance of subentities within the firm, using as inputs the variances generated by the accounting system. Hopefully the arguments to be presented in this paper and the illustration will encourage the designers of managerial information systems to make such analyses an integral part of the accounting systems.