Reviews the book "Scientific Method for Auditing," Lawrence L. Vance.

Auditing, Sampling, Substantive test, Error characteristics

Joseph L. Balintfy, John Neter, William Wasserman, Binary and Chain Comparisons with an Experimental Linear Programming Food Price Index, The Review of Economics and Statistics, Vol. 52, No. 3 (Aug., 1970), pp. 324-330

This article considers audit sampling and the auditor's risks of incorrect conclusions in deciding whether an accounting population contains a material error amount. The proposed new sample evaluation procedure takes into account the correction of all errors observed in the sample and calculates a modified upper bound on the total error amount remaining in the unsampled population. This procedure is known as rectification. It is shown that rectification significantly reduces the risk of incorrect rejection when sample sizes of 150 and 300 are used. Since the size of the error amount observed in the sample is crucial to the extent of risk reduction, the first stage of this study investigated the magnitude of rectification for three dollar-unit sample selection methods commonly used in practice-systematic, random, and cell selection. Theoretical results for the expected error amount contained in systematic samples were obtained for 32 different study populations for three sample sizes (65, 150, and 300). Simulation was used to estimate the expected error amounts with random and cell selection for the same 32 study populations and three sample sizes. The results show that the average error amount subject to rectification with cell selection is very close to that with systematic selection for alt three sample sizes. Random selection produced smaller expected error amounts than did systematic selection, especially for the larger sample sizes. In the second part of the study, the risks of incorrect rejection with and without rectification were compared. Factors considered in this comparison were sample size (65, 150, or 300) and sampling method (systematic, random, or cell). Use of rectification led to noticeable reductions in the risk of incorrect rejection for sample sizes other than 65. For a given application, auditors will need to determine whether the cost of the additional sampling is balanced by the savings from the reductions in the risk of incorrect rejection with rectification.

[This article considers audit sampling and the auditor's risks of incorrect conclusions in deciding whether an accounting population contains a material error amount. The proposed new sample evaluation procedure takes into account the correction of all errors observed in the sample and calculates a modified upper bound on the total error amount remaining in the unsampled population. This procedure is known as rectification. It is shown that rectification significantly reduces the risk of incorrect rejection when sample sizes of 150 and 300 are used. Since the size of the error amount observed in the sample is crucial to the extent of risk reduction, the first stage of this study investigated the magnitude of rectification for three dollar-unit sample selection methods commonly used in practice-systematic, random, and cell selection. Theoretical results for the expected error amount contained in systematic samples were obtained for 32 different study populations for three sample sizes (65, 150, and 300). Simulation was used to estimate the expected error amounts with random and cell selection for the same 32 study populations and three sample sizes. The results show that the average error amount subject to rectification with cell selection is very close to that with systematic selection for all three sample sizes. Random selection produced smaller expected error amounts than did systematic selection, especially for the larger sample sizes. In the second part of the study, the risks of incorrect rejection with and without rectification were compared. Factors considered in this comparison were sample size (65, 150, or 300) and sampling method (systematic, random, or cell). Use of rectification led to noticeable reductions in the risk of incorrect rejection for sample sizes other than 65. For a given application, auditors will need to determine whether the cost of the additional sampling is balanced by the savings from the reductions in the risk of incorrect rejection with rectification.]

ABSTRACT: This note extends the analysis of line-item error taints and error rates originally presented in Johnson, Leitch and Neter [1981] by first considering the distribution of dollar-unit taints, the relevant distribution when simple random sampling is applied to monetary units. Next, empirical evidence on the relation between the taint amount and book amount is presented. Finally, empirical findings on the magnitudes of dollar-unit error rates are provided.

[This note extends the analysis of line-item error taints and error rates originally presented in Johnson, Leitch and Neter [1981] by first considering the distribution of dollar-unit taints, the relevant distribution when simple random sampling is applied to monetary units. Next, empirical evidence on the relation between the taint amount and book amount is presented. Finally, empirical findings on the magnitudes of dollar-unit error rates are provided.]

ABSTRACT: Auditors require empirical information about the characteristics of errors in audit populations. In this paper, the error characteristics in 55 accounts receivable and 26 inventory audits are examined. First, the error rates present in these audits are analyzed, and the balance between overstatement and understatement errors is examined. The distributions of the error amounts and error taintings are then studied, as well as the relation between error amounts and book amounts. Some major findings are contrary to auditors' expectations, emphasizing the need for replication studies to further explore this area.

ABSTRACT: This paper presents a statistical sampling approach based on the multinomial distribution for obtaining a bound for either total population understatement or overstatement errors or both. The approach is nonparametric in nature and, unlike most currently used techniques, it has known characteristics so that the auditor is assured of the specified confidence level regardless of the nature of the population and the nature of the error pattern. Results are presented which show the multinomial bound to give tighter bounds than the Stringer bound in all instances stud]ed. The behavior of the multinomial bound is stud[ed with respect to the effects of sample size and error patterns.

ABSTRACT: A modification of the multinomial bound has been developed which enables the auditor to obtain bounds for substantially larger numbers of errors in audit samples than was possible previously. In this paper, the characteristics of this modified bound are examined and compared with those of the widely used Stringer bound. It is found that the modified multinomial bound is usually considerably tighter than the Stringer bound and should be useful in many audit applications. The confidence level of the modified multinomial bound is also investigated by sampling simulation, and it is found that the level exceeds or is close to the nominal level for all populations studied.