Chapter Fourteen – SamplingSummaryT

Chapter Fourteen – Sampling

Summary

This Chapter explains the basic concepts of sampling and gives specific guidance on the application of sampling when used in substantive testing or when executing tests of controls.
Introduction
14.01 Audit sampling is the application of audit procedures to less than 100% of the population making up an account balance or class of transactions to evaluate evidence as to the overall correctness of account balances or classes of transactions. In Horizon, the term “sampling” implies the use of the Grant Thornton Sampling Plan (GTSP). The components of GTSP are incorporated in Explorer using the sample size calculator.
14.02 The objectives of GTSP are to:
• attain effective and efficient audit samples
• document the considerations that determine sample sizes
• evaluate sample results
• facilitate compliance with professional standards.
14.03 GTSP is a unique approach to audit sampling, a comprehensive and systematic plan that focuses directly on the assessment of audit risk and integrates statistical theory with the use of audit judgment. The plan provides for the translation of subjective audit judgments into quantitative terms for the following key features:
• Materiality and tolerable error
• inherent risk
• control risk
• evidence from other related substantive procedures, including analytical procedures
• expected aggregate error in the population.
14.04 Sometimes, the appropriate response to the risks in an account balance or class of transactions is such that high-value or key-item testing appropriately responds to the risks present.
High-value testing or testing individually significant items is examining 100 percent of the items above a certain monetary unit amount. This is not sampling as the audit team is actually dividing the population (account balance or class of transactions) into two populations, one of which is tested 100% and the other is tested by other means or not tested at all. In this approach, the high-value population that the audit team tests represents a high proportion of the total account balance so that the monetary unit amount of the untested population represents a lower risk of material misstatement. The risk of material misstatement in the untested population can be further reduced by performing other less detailed, substantive procedures such as analytical procedures. Alternatively, the risk of a material error in the untested items may be sufficiently low to warrant no further testing.
Key item testing involves identifying items with qualitative characteristics (unusual, prone to error, or other identified risks), making up a balance or class of transactions that, in the audit team’s judgment should be tested. Key items also may include large dollar items, but may not be individually significant. Again, this is not sampling as the audit team is separating the total population into two smaller populations, high-value and key items. Examples of these qualitative considerations include:
• transactions or balances that are subject to a high degree of management involvement
• transactions or balances that are not in the ordinary course of business, especially with related parties
• balances with unexpectedly low activity for an extended period of time (e.g., slow-moving items in inventories and old outstanding receivables)
• suspicious or unusual items and apparent deviations from the norm
• significant year-end transactions or adjustments.
The selection of key items in a population may be simplified by using IDEA; however, defining what will constitute a key item should be determined by the audit team during risk assessments.
Sampling Risk
14.05 Sampling risk arises from the possibility that a test applied to a sample will result in a conclusion that differs from the conclusion that would be reached if all units in the populations were examined. That is, a particular sample may contain proportionally more or fewer monetary errors than exist in the population. Sampling risk increases from zero as the sample size decreases.
14.06 Two aspects of sampling risk are of concern when performing substantive tests of details:
1. Risk of Incorrect Acceptance is the risk that the sample supports the conclusion that the population is not materially misstated even though, in fact, the population is materially misstated.
2. Risk of Incorrect Rejection is the risk that the sample supports the conclusion that the population is materially misstated even though it is not.
14.07 Sampling risk is determined and controlled in a statistical sample. Although sampling risk is not quantified for a nonstatistical sample, the extent of such risk is reduced by representative selection and by stratification.
Nonsampling Risk
14.08 Nonsampling risk includes all aspects of audit risk that are not due to sampling. Nonsampling risk includes the possibility of selecting audit procedures that are not appropriate for the audit objective (e.g., confirming recorded receivables cannot be expected to detect unrecorded receivables). Nonsampling risk also includes the possibility that errors in documents examined may not be recognized. This failure will make the procedure ineffective, even if all items are examined.
Plan the Test
14.09 To properly plan a sampling test, the audit team should:
• Decide whether sampling is the correct strategy
• Understand sampling test objective
• Define what is an error
• Determine the population to test
• Identify individually significant and other key items in the population
• Define sampling unit
Decide Whether Sampling is the Correct Strategy
14.10 Horizon does not require sampling, but it is suggested as the testing strategy for a number of critical assertions. The audit team should decide whether sampling is the most effective or cost-effective audit approach. Conditions to be considered in deciding whether to use sampling include:
1. Whether the contemplated audit procedures are effective when applied on a test basis - Some audit procedures (e.g., analytical procedures or testing unusual or individually significant items) are not generally applied to a sample. For example, because the power of analytical procedures is derived from analyzing all the data and their relationships, sampling is not generally used.
2. The cost-benefit relationship - Audit procedures that provide a relatively high degree of evidence at relatively low cost are frequently applied on a more extensive basis than a representative sample. Sampling should generally be avoided whenever more cost-effective procedures are available (e.g., significant coverage by examining large monetary values or unusual items based on quantitative or qualitative characteristics, applying analytical procedures, or a combination of both such approaches). IDEA is an excellent tool for analyzing populations and identifying unusual items. For example, we might recompute depreciation expense 100% with IDEA, since extensive evidence is provided at relatively low audit cost.
3. The potential for material error – Horizon does not recommend sampling in situations when the risk of potential errors is immaterial for a specific assertion. These assertions are usually considered not critical (see the discussion of “D” procedures in Chapter 1) and are generally audited more effectively by scanning the account detail to identify items for testing or utilizing analytical procedures.
14.11 Sampling is ordinarily preferred for a population made up of a large number of individually insignificant amounts that, in the aggregate, are material to the financial statements (e.g., trade accounts receivable). This is especially true if the account in question is a balance sheet account. Additional factors that might indicate that sampling is appropriate are:
• the potential risk of a material misstatement in a population is not low
• internal control weaknesses exist
• it is necessary to develop an estimate of the account balance or class of transactions instead of testing an existing balance (e.g., determining the value of inventory because the inventory prices for individual items are not updated throughout the year).
14.12 Sampling may be used as a substantive test or for testing key controls, depending on whether the audit team is attempting to evaluate an amount in a balance or class of transactions or the operating effectiveness of identified key controls.
Understand Sampling Test Objective
14.13 The objective of a sampling procedure differs depending on whether the procedure is a substantive procedure or a test of controls. The objective of a substantive procedure is to gather evidence about an assertion. The objective of a test of controls is to gather evidence relating to the operating effectiveness of internal controls. Sample design is a function of test objective. Therefore, when sampling the audit team should identify the objective of the test in order to determine the appropriate sample design. Ordinarily, samples designed as tests of controls do not provide sufficient evidence to reach a meaningful monetary (substantive) conclusion.
14.14 The following are examples of test objectives:
Test Objective
Confirmation of receivables Determine whether the accounts receivable exist and the sales occurred
Inventory test counts Determine whether the quantities are accurate and exist
Inventory price tests Determine whether the recorded amount is correct (Is it priced correctly?)
Define What is an Error
Substantive Samples
14.15 Errors should be defined by the audit team during the design of the sampling application. An error (intentional or unintentional) is a matter that causes a misstatement in the financial statements and can involve:
• unintentional mistakes or omission of amounts in an account balance or class of transactions, including mathematical or clerical mistakes in the underlying records and accounting data from which the financial statements were prepared
• mistakes in t