Your audit firm, Garrett and Schulzke LLP, is engaged to perform the annual audit of Hooplah, Inc., for the year ending December 31, 2017. Hooplah is a privately-held company that sells electronics components to companies that manufacture various appliances. The company hires a public accounting firm to provide an audit of its financial statements in order to get favorable terms on its bank loans. Your firm has audited Hooplah for the past three years. For the current audit engagement, your team has already performed most of the audit work; however, there are a few loose ends for you to tie up. Portions of Garrett and Schulzke’s audit policy relating to audit sampling are provided to assist you in completing the procedures.
Audit sampling is commonly applied in performing tests of controls and tests of details. Audit sampling involves the application of audit procedures to less than 100 percent of the items in a population of audit relevance, selected in such a way that the auditor expects the sample to be representative of the population and thus likely to provide a reasonable basis for conclusions about the population. A sample is usually selected either randomly (using some form of random-number generator) or haphazardly (where the auditor attempts to select items randomly but without using a formal random-number generator).
Auditors often use sampling approaches that involve formal statistical theories and principles, similar to those you may have learned in an introductory statistics class. Statistical sampling applications require the use of random selection, based on a formal random-number generator (such as the one built in to Microsoft Excel or audit software such as ACL). Auditing standards also allow auditors to use “non-statistical” sampling approaches. Non-statistical audit sampling approaches are based on a foundation of statistical principles, but allow certain departures from formal statistics in order to simplify the auditor’s task. One such simplification is that non- statistical sampling allows the use of haphazard selection of the items to be examined.
When an auditor examines only a sample instead of all of the items in the population, an element of uncertainty enters into the auditor’s conclusions. This uncertainty, referred to as sampling risk, is due to the possibility that the sample selected is not representative of the population and that as a result the auditor will reach an incorrect conclusion about the population. It is crucial that sampling risk be taken into account when evaluating the results of any audit procedures that involve sampling.
Sampling approaches also differ depending on the nature of the items the auditor is examining and the objectives of the auditor. Depending on the method of sampling used, different formulas and tables are available to help you determine the number of items to include in the sample.This case has two parts. In Part A you will be asked to use statistical attribute sampling for testing controls. In Part B you will use non-statistical substantive sampling for testing accounts receivable.
PART A Tests of Controls
Before beginning substantive tests for accounts receivable, you and Darrell need to perform tests of important controls over the revenue process to justify the preliminary control risk assessment of low control risk and to determine whether a reliance approach is appropriate for substantive testing. You have decided to test three controls in Hooplah’s revenue process:
1. Sales are properly authorized for credit approval.
2. Sales are reviewed by the sales manager to ensure that they are properly priced.
3. Credit memos for sales returns are properly authorized once all goods have been returned.
Garrett and Schulzke’s sampling guidance for testing controls is based on attribute sampling. In testing the operating effectiveness of controls you are interested in determining the likely “deviation rate,” or the rate at which a control is not properly operating. Testing controls to measure the deviation rate provides evidence about whether the control is operating effectively an acceptably high percentage of the time and helps the auditor determine whether or not the control can be relied upon. Garrett and Schulzke’s sample size table for attribute sampling can be found in Appendix A. Use the table to determine the appropriate sample size for tests of controls given a specified tolerable deviation rate and estimated deviation rate. Garrett and Schulzke’s controls testing sampling policy indicates that to place high reliance on controls (i.e., to support a low level of remaining control risk) the test must be performed at a high level of assurance, which they define as 95 percent confidence. Thus, only the sample size table associated with 95 percent confidence is provided. In addition to the sample size table, an evaluation table is provided in Appendix A that will help you determine the computed upper deviation rate for the control given the number of detected deviations and the sample size used. The computed upper deviation rate is the sum of the sample deviation rate and an appropriate allowance for sampling risk. In other words, it represents the upper end of the 95 percent confidence interval for the deviation rate in the population. If the computed upper deviation rate indicated is greater than the tolerable deviation rate for the control, the auditor should not rely on the control.
In order to determine the appropriate sample size for your tests, the importance of the control to be tested needs to be assessed. Garrett and Schulzke’s policy indicates that any controls the auditor might consider relying on are either “highly important” or “moderately important.” According to the firm’s policies, a deviation rate of only four percent can be tolerated for controls deemed “highly important,” while a tolerable deviation rate of eight percent can be tolerated for controls deemed “moderately important.” Some “tolerance” for error must always exist when using audit sampling because sampling involves sampling risk. The higher the tolerable deviation rate, the lower the sample size.
Based on your knowledge of Hoopla’s systems, you decide that the first two controls are moderately important, and that the third control is highly important. The other input needed to determine sample size is the estimated population deviation rate. Based on past experience with the client and considering historical rates, you determine that the estimated population deviation rates for the three controls are one percent, two percent, and zero percent, respectively.1
Darrell decided that since all sales pass through the first two controls and since the tolerable deviation rate for those controls is the same, he could be more efficient by using the same sample of transactions to test both controls. He randomly chose 58 sales transactions and tested the documentation for evidence of proper credit approval and review of pricing by the sales manager. After finding only one exception for the first control and two exceptions for the second control, Darrell concluded that both controls are operating effectively.
Now that Darrell has tested the first two controls, you need to test the third one before moving on to substantive testing.
REQUIRED  Evaluate the appropriateness of Darrell’s conclusions relating to the first two controls:
[a] Is it acceptable to use the same set of transactions and the same sample size to test two different controls?
[b] Do you agree with Darrell’s conclusions with respect to these first two controls? If not, why not? In evaluating Darrell’s conclusions, you may wish to refer to the attribute sampling evaluation table available in Appendix A.
[c] What additional work, if any, is necessary to support his assessment that both controls are operating effectively?
 Evaluate each of the following questions independently:
[a] Referring to Garrett and Schulzke’s sample size table (Appendix A), determine an appropriate sample size for the test of the control relating to authorization of credit memos.
[b] Regardless of your answer to question [a], assume that Darrell randomly selected a sample of 75 credit memo packages, each of which contains a credit memo with a matching receiving report and inventory warehouse receipt, and has reviewed all but the last five credit memo packages and found no exceptions. Each credit memo authorizing a customer refund or reduction in the amount owed due to the return of goods should be authorized by Brian Thompson, the accounting supervisor over accounts receivable. When a customer wants to return unwanted or defective product, they go to Hooplah’s website and download a “Customer Return Report” and fill in their information as well as a description and quantity of goods being returned. Chris Jacobs in the receiving department at Hooplah uses the Customer Return Report as the receiving report when the shipment comes in, and Felix Katt counts and inspects the goods to make sure they’re all in good condition.The goods are then transferred back to Jed Baxter in the warehouse, who issues an “Inventory Receipt.” Once the Inventory Receipt is attached to the Customer Return Report, the documents are forwarded to Brian so that he can approve a credit memo. Brian examines the Customer Return Report and the Inventory Receipt to ensure that credit is only authorized when goods have been received and for the quantity actually received. He then documents his authorization by marking his initials on the credit memo. You have agreed to help Darrell by examining the remaining five credit memo packages. You will find the last five credit memo packages at www.pearsonhighered.com/beasley. Examine these documents for Brian Thompson’s initials indicating the memos were approved (for this question, just focus on his initials for audit evidence that the control is operating effectively) and then evaluate the test results for the full sample. Provide support for your assessment on the effectiveness of the control based on testing performed.
 In question 2[b] you gathered evidence on the operation of the control by examining the five credit memos for Brian’s initials. While this provides some evidence of control, higher quality evidence that the control is actually operating effectively can be obtained by reperforming Brian’s control procedure. This would be done by verifying that each credit memo is supported by a Customer Return Report and Inventory Receipt and that the quantity and description of goods on the credit memo is supported by the quantity and description on the supporting documents. If Brian’s initials are on the credit memo, but the quantity or description of goods on the memo is not consistent with the quantity and description on the supporting documents, this would be considered a control deviation. Assume Darrell reperformed the control for the first 70 credit memo packages and found no exceptions.
[a] Reperform the control for the remaining five credit memo packages and evaluate the test results for the full sample. Provide support for your assessment on the effectiveness of the control based on testing performed.
[b] If you came to a different conclusion in 2[b] and 3[a], which conclusion is more supportable and why?
 Assuming the controls testing is not expanded to provide additional support regarding the effectiveness of the controls tested, what are the implications of the controls testing in question 3[a] with respect to the nature, timing, and extent of substantive evidence that must be gathered to support the fairness of the accounts receivable balance?
Tests of Details PART B
Regardless of what you found in Part A, for Part B assume that you are able to place moderate reliance on the controls tested and that you have already obtained some substantive evidence supporting the fairness of accounts receivable from substantive analytical procedures. While you have already obtained some assurance regarding the fairness of the ending accounts receivable balance, you do not yet have sufficient evidence given the size of accounts receivable and the remaining risk of misstatement. You plan to request that some of Hooplah’s customers confirm their accounts receivable balance directly to you. In the prior year’s audit, aggregate misstatements of less than 0.5% of the accounts receivable balance were discovered via customer confirmation testing. The few misstatements that were found were promptly corrected by Hooplah.
The current-year information that follows will help you in determining the nature and extent of detail testing in order to have sufficient appropriate evidence to conclude on the fairness of the accounts receivable balance.
Current Year Information:
Net income = $9 million Total assets = $85 million
Total accounts receivable = $12,881,551
Accounts receivable greater than 90 days past due = $2 million Tolerable misstatement for accounts receivable = $400,000
In most cases, the selection of items to be detail tested is based on two approaches, which can be used singly or in combination to achieve the desired level of assurance with respect to the population being tested:
1 - Directed Testing 2 - Audit sampling
Directed testing, also known as “targeted testing” or “key item testing,” is a technique that involves selecting items to examine based on a particular characteristic of interest such as size or risk. Unlike audit sampling, the items are not randomly (or haphazardly) selected. Instead, selection is “directed” or “targeted” based on a particular characteristic.Thus, directed testing is not considered “sampling” per se, because the subset of selected items is not expected to be representative of the population. Garrett and Schulzke’s audit policy requires that teams direct test all individual items in the account that are greater than tolerable misstatement. Thus, even if the auditor intends to perform audit sampling to test an account (e.g., accounts receivable), the auditor must first examine all items (e.g., individual customer accounts) that are individually greater than tolerable misstatement.
After testing all such items, it is often appropriate to expand the directed testing to specifically select relatively high risk items, if such items can be identified. The auditor may also expand directed testing to select relatively large items other than those that are larger than tolerable misstatement in order to achieve “coverage” of a higher dollar percentage of the total account. Selection criteria for directed testing can include a combination of risk and size components. Expanding the number of items examined in directed testing can often provide sufficient assurance in combination with the assurance already obtained from other audit procedures (e.g., risk assessment, controls testing, substantive analytical procedures, testing in related accounts, etc.). In such cases, the use of an audit sampling approach is unnecessary.
Garrett and Schulzke’s substantive audit sampling policy uses a nonstatistical sampling approach. Items are selected from the population either randomly or haphazardly, at the auditor’s discretion. To determine the appropriate sample size, the firm provides the following formula:
The sampling population book value is the total book value of all the items available to be selected in the sample.This total does not include items already removed for direct testing (i.e., all items greater than tolerable misstatement and other items selected based on size and/or risk characteristics).Tolerable misstatement is the greatest amount of misstatement that can be tolerated for the account being tested without concluding that the account is materially misstated. Expected misstatement is the amount of misstatement that the auditor expects to find in the account being tested.The “confidence factor” included in the above equation is determined based on the assessed risk of material misstatement for the account and the desired level of confidence from the sample. The confidence factor table below is from Garrett and Schulzke’s sampling policy.
The purpose of audit sampling is to draw conclusions about the entire population through testing a subset of the population. To draw inferences about the entire population, sample results must be projected to the population. Garrett and Schulzke’s sampling policy provides two projection methods: ratio projection and difference projection. Ratio projection is performed by calculating the ratio of the misstatement to the sample book value and projecting it to the sampling population book value according to the following formula:
Difference projection is performed by calculating the average misstatement per sample item (e.g. individual customer account) and projecting it to the number of items in the sampling population according to the following formula:
REQUIRED  In selecting which customer balances to detail test via accounts receivable confirmations, assume that you have decided to direct test only the minimum number of customer accounts required; that is, you will direct test only customer accounts that are greater than tolerable misstatement, and you will use audit sampling to test the remainder of the population. You can find the accounts receivable detail listing at www.pearsonhighered.com/beasley. As noted there, Hooplah has a total of 357 customers, with an accounts receivable balance totaling $12,881,551. Based on the engagement team’s knowledge of Hooplah’s accounts receivable processes and policies, experience in prior year’s audits, and the results of tests of controls and substantive analytical procedures, the assessment of risk of material misstatement for accounts receivable has been set at “moderate.” Prepare a schedule that includes the following [a] List the customer number and related balance for all customers you plan to direct test. [b] Indicate your computed sample size (using the sample size formula provided on prior page). Provide supporting calculations and justification for your sample size, including justification for the confidence factor and the level of expected misstatement you used to compute your sample size.  Based on the same background information as was used for question 1, but assuming that in selecting which customer balances to detail test you want to expand directed testing by selecting additional items based on risk and size, reevaluate the mix of directed testing and audit sampling. If you believe it would be efficient and effective to increase your directed testing, prepare a schedule that includes the following: [a] Identify what characteristic can be used to select riskier items. [b] List the customer numbers and related balances you would select for directed testing based on risk and provide the characteristics you used. [c] List the additional customer accounts you would select for directed testing based on size and “coverage.” [d] Determine whether it would be necessary to test the remaining population using audit sampling; if so, compute your sample size for testing the remaining population through audit sampling and justify the inputs you used in the sample size formula.  Which detail testing approach seems most appropriate in this situation: the minimum level of directed testing together with a larger audit sample, expanded directed testing with no audit sampling, or both expanded directed testing and audit sampling? Be sure to consider the effectiveness and efficiency of the approach, as well as the level of assurance needed in view of the evidence already obtained from controls testing, substantive analytical procedures, etc.  Independent of your responses to prior questions, assume that you direct tested customer balances greater than tolerable misstatement and randomly selected a sample of 40 additional customer balances for confirmation. The total book value of the 40 items sampled is $761,030. No differences were noted in the directed testing, and the sample yielded a combined overstatement in Hooplah’s records of $4,215. Brian Thompson, the accounts receivable supervisor agrees that the differences noted are misstatements due to pricing errors. Please answer the following questions: [a] How much is the known misstatement in the accounts receivable balance? [b] How much is the projected misstatement in the population (i.e., the total accounts receivable account) using ratio projection? [c] How much is the projected misstatement in the population using difference projection? [d] Explain why the two projections produce different results and describe the circumstances under which one projection approach might be more appropriate than the other. [e] Based on the results of the detail testing outlined in this requirement, as well as the assurance obtained from controls testing and substantive analytical procedures, do the audit procedures support the assertion that the accounts receivable account is fairly stated? Why or why not? PROFESSIONAL JUDGMENT QUESTIONS It is recommended that you read the Professional Judgment Introduction found at the beginning of this book prior to responding to the following questions.  How might the availability tendency contribute to less than optimal auditor judgment in the sampling process? How might an auditor mitigate the effects of this tendency?  How might the confirmation tendency contribute to less than optimal auditor judgment in the sampling process? How might an auditor mitigate the effects of this tendency?