Knowledge that Transforms

Abstract The article reports that the use of the entropy measure must be a function of its context and that the meaning attributable to the entropy should be determined by the nature of the input data which are used in computing the measure. Thus, the same measure of the entropy could be used to measure uncertainty, ignorance, or information, and to measure decomposition. The latter is obtained when proportions are substituted for probabilities. The results of this study also suggest that the indiscriminant utilization of the entropy in accounting measurements should be a matter of concern. Finally, there is a great need in accounting to measure the level of data aggregation which would be considered optimal. Although the results of this research did not support the entropy for fulfilling such a need, they did not preclude the possibility of the usefulness of the measure given the appropriate formulation and decision situation. The entropy may only be used to make a choice from among competing sets, which are acceptable methods of aggregation under the prevailing accounting conventions.