Attribute sampling is used to analyze the characteristics of a population and also helps in determining whether a company’s internal controls are being followed. Controls data is essential for quality auditing and cost reduction (Westland, 2020). In this situation, it is important for the auditor to choose the correct sample sizes and structure in order to avoid audit errors. The main reason for the occurrence of such errors is the choice of sample, which does not reflect the characteristics of the entire population (Westland, 2020). The auditor in the case presented made a number of sampling errors that led to incorrect conclusions and judgments.
First of all, statistical sampling does not eliminate the need for judgment. Although in contrast to non-statistical sampling, this technique gives each unit an equal chance of being selected. However, the choice of statistical sampling should be based on population characteristics and audit objectives. In particular, it is necessary to consider which sampling is appropriate: attribute-based or variable-based. The tolerable rate of deviation of 20% is also too high for the population. To select an appropriate rate, it is necessary to consider deviations in the population, as well as an allowance for sampling risk. Baker also deferred the consideration of the allowable risk of overreliance, which is an incorrect assumption. This attribute must be set immediately after the tolerable rate of deviation. Consideration of this value is important for correctly assessing the results of the audit, as well as making adjustments to the sample size.
The main auditor’s mistake, in this case, is randomly selected. This approach is not consistent with the statistical sampling technique chosen by the auditor. First of all, the sample size was too small to obtain reliable audit results. Additionally, having identified an error in the invoice, the auditor did not consider it as a necessary deviation and did not take it into account when analyzing the audit results. It is also important that the allowances for sample risk were not calculated correctly. To calculate this attribute, the auditor needed to consider such indicators as tolerable deviation, as well as the expected mean of the population. Since the sample was chosen randomly for the population, the last indicator was not determined, which does not allow us to draw correct conclusions.
Additionally, the population from which the sample selection was made was incorrect and could not be used for auditing. First of all, the auditor used an extremely small sample, which could not represent the characteristics of the entire population. Additionally, the sample was saddled randomly rather than based on similarity in population characteristics, which also makes it invalid. Although increasing the number of sample items does not have a significant effect on audit results, the selection of additional items was also random. This approach distorts the conclusions that can be drawn on the basis of such a sample.
Systematic sampling, which was chosen as the main technique, in this case, is appropriate. However, systematic sampling uses the selection of elements with a set frequency, which was not done in this case. The auditor did not use systematic sampling but random sampling, which makes it invalid. Thus, in the presented case, the choice of sample size and composition was not correct, which led to erroneous judgments as a result. Thus, the statement that the sample supported a low level of control risk is not true.
References
Westland, J. C. (2020). Audit analytics: Data science for the accounting profession. Springer International Publishing.