Fang Kuo Hu

Publication Date


Document Type


First Advisor

Cheng, Philip C., 1925-

Degree Name

M.B.A. (Master of Business Administration)

Legacy Department

Department of Accountancy


Cost accounting


In these days, the predicting of cost for managerial planning has been emphasized more than ever before. Thus, the estimating of future cost is one of the accountants' indispensable responsibilities. In the short run, while the production process, organization structure and plant facilities can be held constant; the estimating of future cost involves the past cost behavior analysis. Under conditions of certainty, the cost behavior in the past can be used as a model to predict future cost. Among approaches to the determination of cost behavior, the use of statistical regression affords the most rigorous framework within which the cost function can be estimated. The choice of regression equation is based on the assumption of cost behavior. Accountants assume the linear relationship between cost and output while economists assume the nonlinear relationship between them. Both types of cost behavior should be examined in order to justify the selection of linear regression equation. Under the assumption of linear relationship between cost and independent variables, the simple linear regression and multiple linear regression can be used for the cost analysis and future cost prediction. Both regression techniques have their merits and demerits. A clear understanding of their feasible conditions is necessary in order to avoid the misuse of either regression technique. An empirical research was used to test the theoretical support for both regression techniques. The results of analyzing time serial data show that both regression techniques have their feasible conditions. In some cases the situation is suitable for the application of both methods. The choice of regression equations depends upon the one which can yield a smaller standard error of estimating a dependent variable. In the case where the standard error of estimating regression coefficients is too large, neither the simple linear regression technique nor the multiple linear regression technique can be used to depict cost behavior or to predict future cost even under the conditions of certainty.


Includes bibliographical references.||Includes illustrations.


vii, 99 pages




Northern Illinois University

Rights Statement

In Copyright

Rights Statement 2

NIU theses are protected by copyright. They may be viewed from Huskie Commons for any purpose, but reproduction or distribution in any format is prohibited without the written permission of the authors.

Media Type