Abstract The article presets a reply from William Steve Albrecht on comments made by David E. Keys and Curtis Norton on the topic related to Estimation Error in Income Determination. Certainly, their critique clarifies many of the implicit assumptions and ambiguities in the original paper. The author feels that there are three comments made by Keys and Norton that deserve clarification. They suggest that the author implied the relative frequency interpretation of probabilities that will always be preferable to the subjective interpretation of probabilities in determining forecast error. Since they are essentially the same, the author agreed that a decision about the preferability of one interpretation over the other is a cost-benefit question and is probably situation-specific. Keys and Norton imply that it may have been inappropriate to use a quantification process in the time domain because the underlying process of each time period may not be constant or stable. They also suggested that the level of aggregation of the variances and covariance can affect the magnitude of the confidence interval around net income.

Abstract This article focuses on the problem of estimation error in income determination. The author aims to provide a framework within which the uncertainty of the income statement can be computed and interpreted and then to illustrate, using a case study, the degree of uncertainty and resulting imprecision that does exist in one firm's earning computation. In accounting, two variables, suggestively, are of considerable interest: first, the flow of earnings over a specified period, and second, the stock of wealth at the end of that period. Based on the study, it is concluded that forecasting error can arise from two sources. There can be forecasting error about the amount of an item, such as, salvage value of fixed assets or the amount of uncollectible accounts; or there can be forecasting error about the timing of an item, such as, service lives of fixed assets. Another source of uncertainty in financial statements results from measurement or counting errors. Hence, an artificial partitioning of the entity's life to obtain intermediate progress data regarding enterprise goal accomplishment is needed. The quantification of uncertainty in a period's earnings computation is a two-step process. The first step involves assessing the estimation error inherent in each of the determinants of net income; the second step involves aggregating the uncertainty variables to find the cumulative effect for the period.

Abstract Comments on an article about the accuracy of corporate and security analyst's forecast of earnings. Conditions for accurate forecasting; Criticisms on the article.