Evaluating Bias and Imprecision of Costs Objects in Sophisticated Costing Systems
51 Pages Posted: 18 Aug 2018 Last revised: 8 Dec 2018
Date Written: December 7, 2018
Bias and imprecision are crucial quality dimensions of measurement and therefore of accounting information as well. This paper is the first to investigate imprecision – and to a lesser extent also bias – of cost objects via numerical experiments. With respect to activity-based costing systems, we show that one must expect within a portfolio a substantial number of cost objects with low precision. This is even the case, when only a few random measurement errors are given (‘bad seed effect’). We document how imprecision can increase considerably error in single measurements, as it adds up with bias. We also analyze a number of drivers of bias and imprecision and identify several recurrent patterns at the cost object level, providing rules of thumb concerning cost error behavior. Our results suggest an extended costing error picture at the cost object level and show that bias and imprecision have different implications for the usability of cost information.
Keywords: cost accounting, error, bias, imprecision, information content
Suggested Citation: Suggested Citation