Audit Evidence and Sampling — Practice

Question 1

An auditor tests 500 low-value, routine supplier invoices for a balance where the real risk lies in a handful of large, complex, judgement-heavy year-end accruals that were not included in the sample at all. Explain, using the concepts of sufficiency and appropriateness, what has gone wrong.

Accepted Answer

This is a failure of appropriateness masquerading as an apparently adequate quantity of evidence, and it illustrates precisely why sufficiency and appropriateness must both be satisfied and neither substitutes for the other. Sufficiency, the quantity of evidence, might appear satisfied here, since 500 invoices is a substantial sample by volume; but appropriateness, the relevance and reliability of that evidence to the actual risk present, has not been satisfied at all, because the real risk in this balance lies specifically in a handful of large, complex, judgement-heavy year-end accruals, and testing 500 routine, low-value, presumably lower-risk supplier invoices provides evidence about a part of the population that was never where the risk actually was. A large volume of evidence about the wrong part of a population, or about a part carrying comparatively little risk, does not compensate for the absence of evidence about the part that genuinely carries the risk; no amount of additional routine-invoice testing would ever provide relevant evidence about whether the large, judgemental year-end accruals are appropriately stated, since routine invoices and complex accruals are simply different things testing one does not illuminate the other. The auditor's sample design has failed to be risk-responsive: audit procedures are meant to be designed based on the assessed risk of material misstatement, concentrating evidence-gathering effort where risk is genuinely concentrated, and a sample that is large in volume but misdirected relative to where the actual risk sits produces evidence that is insufficient in the appropriateness sense, however large it is in the sufficiency sense.

Question 2

Rank the following four pieces of evidence from most reliable to least reliable, applying the generalisations in this chapter: (a) an oral assurance from the client's accountant that a reconciling item is immaterial; (b) a written confirmation received directly from the client's bank; (c) a photocopy of a signed lease agreement provided by the client; (d) the auditor's own observation of a year-end inventory count.

Accepted Answer

Applying the reliability generalisations in order of decreasing reliability: item (b), the written bank confirmation, ranks most reliable, since it is evidence obtained from an independent source outside the entity, in documentary form, and represents a direct written response to the auditor specifically, all features the hierarchy associates with higher reliability. Item (d), the auditor's own observation of the year-end inventory count, ranks next, since evidence obtained directly by the auditor is more reliable than evidence obtained indirectly or by inference, and direct observation of a physical count is a strong, first-hand form of evidence, though it is limited to the specific point in time observed and does not itself confirm valuation or ownership, only physical existence at the time of the count. Item (c), the photocopy of the signed lease agreement, ranks below both, since evidence provided by original documents is more reliable than that provided by photocopies or facsimiles, a photocopy carrying some risk of alteration or incompleteness that an original does not, even though it is still documentary evidence and therefore more reliable than a purely oral statement. Item (a), the oral assurance from the client's accountant, ranks least reliable of the four, since it is both an internal source, connected to the entity being audited rather than independent of it, and oral rather than documentary, combining two separate factors the hierarchy identifies as reducing reliability, and oral assurances of this kind are generally treated as the weakest standalone evidence, warranting corroboration through other procedures before being relied upon.

Question 3

For each identified risk, state the primary assertion at risk and a procedure that would address it: (a) risk that recorded receivables include amounts from customers who never actually placed the underlying orders; (b) risk that a provision for warranty claims is understated relative to the entity's actual future obligation; (c) risk that some genuine purchases made just before year end were recorded in the following period instead.

Accepted Answer

Item (a) is primarily an occurrence risk, since occurrence, a transaction assertion, concerns whether recorded transactions have actually occurred and relate to the entity; recording receivables from customers who never placed the underlying orders means transactions have been recorded that did not, in substance, actually occur. An appropriate procedure is external confirmation, sending confirmation requests directly to the customers whose balances are recorded, since a genuine customer would confirm the balance while a fictitious or non-existent transaction would either receive no response or a response denying the balance, directly testing whether the underlying transaction genuinely occurred. Item (b) is primarily an accuracy, valuation and allocation risk, a balance assertion concerned with whether amounts, including valuation adjustments such as a provision, are appropriately recorded; an understated warranty provision means the recorded amount does not reflect an appropriate valuation of the entity's actual obligation. An appropriate procedure is recalculation and analytical review of the provision, examining the entity's own methodology and historical claims experience, comparing the provision against actual claims paid in recent periods and against industry benchmarks, and assessing whether management's underlying assumptions about claim rates and average claim cost are reasonable and consistently applied, since valuation risks of this kind are addressed by testing the basis and calculation of an estimate rather than by confirming the physical existence of anything. Item (c) is a cutoff risk, a transaction assertion concerned specifically with whether transactions have been recorded in the correct accounting period; genuine purchases made before year end but recorded in the following period is exactly a cutoff error. An appropriate procedure is a cutoff test, examining goods received notes and purchase invoices dated around the year-end date and tracing them to the accounting records to confirm they were recorded in the period the underlying goods receipt actually occurred, rather than in whatever later period the invoice happened to be processed.

Question 4

Name the seven types of audit procedure for obtaining evidence described in this chapter, and explain why inquiry, though listed among them, is described as the weakest form of evidence standing alone.

Accepted Answer

The seven types are inspection, examining records, documents or physical assets; observation, watching a process or procedure being performed; external confirmation, a direct written response obtained from a third party; recalculation, checking the mathematical accuracy of documents or records; reperformance, the auditor's own independent execution of procedures or controls originally performed by the entity; analytical procedures, evaluating financial information through analysis of plausible relationships among data; and inquiry, seeking information from knowledgeable persons within or outside the entity. Inquiry is described as the weakest form of evidence standing alone because it depends entirely on the knowledge, candour and objectivity of the person being asked, none of which the auditor can independently verify merely from the fact that an answer was given confidently or by someone in a position of apparent authority; the person responding may genuinely not know the answer despite believing they do, may unintentionally give an incomplete or imprecise response, or in cases involving fraud may deliberately provide a false or misleading answer to conceal wrongdoing. Unlike inspection of a document, which the auditor can directly examine and assess, or external confirmation, which comes from an independent source outside the entity's control, inquiry's reliability is entirely contingent on the honesty and accuracy of the specific individual responding, over which the auditor has no independent check at the moment the inquiry is made. For this reason, inquiry is almost always corroborated with other procedures rather than relied upon in isolation to support a significant conclusion, though it remains a valuable and necessary procedure for understanding management's views, explanations and intentions, which no other procedure can directly access.

Question 5

Explain the two directions of sampling risk for a test of details, name which is more serious and why, and describe the practical consequence of an auditor who is excessively cautious about the less serious direction.

Accepted Answer

For a test of details, sampling risk runs in two directions: the risk of incorrect acceptance, concluding a balance is not materially misstated when in fact it is, and the risk of incorrect rejection, concluding a balance is materially misstated when in fact it is not. Incorrect acceptance is the more serious of the two, because it directly threatens audit effectiveness: if the auditor incorrectly accepts a balance that is, in truth, materially misstated, the audit may proceed to an inappropriate opinion, expressing an unqualified view on financial statements that in fact contain a material misstatement, which is exactly the failure the whole audit process exists to prevent and which carries the most serious professional and legal consequences for the auditor. Incorrect rejection, by contrast, threatens only audit efficiency: the auditor concludes, wrongly, that a balance is materially misstated when it is not, and this leads to unnecessary further work, additional procedures, further inquiry, perhaps escalation to more senior team members, all aimed at resolving an apparent problem that does not actually exist, consuming time and cost but not resulting in an inappropriate opinion, since the additional work would eventually reveal that the balance was in fact fairly stated. An auditor excessively cautious about incorrect rejection specifically, to the point of biasing sample design or evaluation to avoid it, would in practice be accepting a higher risk of incorrect acceptance instead, since the two risks trade off against each other for a given sample design and cannot both be minimised simultaneously without simply increasing the sample size; this is precisely why properly designed audit sampling approaches are, if anything, deliberately biased towards tolerating the efficiency-costing error of incorrect rejection over the effectiveness-costing error of incorrect acceptance, accepting some risk of unnecessary extra work as the price of protecting against the far more serious risk of missing a genuine material misstatement.

Question 6

Distinguish sampling risk from non-sampling risk, and explain why increasing sample size addresses one but not the other, giving an example of a non-sampling risk that a larger sample size would not fix.

Accepted Answer

Sampling risk is the risk that the auditor's conclusion based on a sample may differ from the conclusion that would be reached if the entire population were subjected to the same audit procedure; it is an inherent consequence of testing less than the whole population, and it exists purely because a sample, however carefully selected, is not the same as the whole. Non-sampling risk is the risk that the auditor reaches an erroneous conclusion for any reason not related to sampling risk, such as applying audit procedures that are not appropriate to the objective of the test, or failing to recognise a misstatement in evidence that was in fact examined; it is a risk of human or procedural error in the conduct of the audit rather than a consequence of the mathematics of sampling. Increasing sample size addresses sampling risk because, as more of the population is examined, the sample increasingly resembles the whole population, and the probability that the sample's characteristics diverge materially from the population's true characteristics correspondingly falls, converging towards zero sampling risk if the entire population were tested. Increasing sample size does not address non-sampling risk at all, because non-sampling risk arises from errors in judgement, technique or attention rather than from the incompleteness of the sample; testing more items using the same flawed procedure, or with the same auditor failing to recognise a misstatement when it appears in the evidence, simply repeats the same underlying error more times rather than correcting it. An example of a non-sampling risk a larger sample would not fix: an auditor who does not understand what a properly authorised purchase order should look like for this client, and therefore fails to notice that an unauthorised purchase order is missing a required approval signature, would make the identical mistake whether examining 50 purchase orders or 500, since the flaw lies in the auditor's own knowledge and attention, not in how much of the population was sampled; this kind of risk is instead addressed through proper training, supervision, and review of the work performed.

Question 7

Distinguish statistical sampling from non-statistical sampling, and explain why neither is described as inherently superior to the other.

Accepted Answer

Statistical sampling is an approach to sampling that uses random selection of the sample items, and applies the laws of probability to evaluate the results, including measuring sampling risk in quantifiable terms; it allows the auditor to make statements about the population with a stated, mathematically derived level of confidence. Non-statistical sampling does not use these techniques; it relies instead on the auditor's own professional judgement, both in selecting the sample, which need not be random, and in evaluating the results obtained, without the formal probability-based measurement of sampling risk that statistical sampling provides. Neither approach is described as inherently superior to the other because both, properly applied, can provide the auditor with a reasonable basis for drawing conclusions about the population; the choice between them is a matter of the auditor's professional judgement about which approach is more suitable and cost-effective in the specific circumstances of a particular audit area, rather than a requirement that one is invariably the correct method. Statistical sampling offers the specific advantage of an objective, quantifiable measure of sampling risk, which can be valuable in supporting the audit conclusion and in documentation, particularly for larger, more homogeneous populations where the underlying statistical assumptions hold reasonably well; non-statistical sampling can be more flexible and can draw more directly on the auditor's accumulated experience and judgement about a specific client or area, which may be particularly valuable where the population is smaller, less homogeneous, or where qualitative factors the auditor has observed are difficult to incorporate into a purely statistical framework. Both remain fully acceptable approaches under the Standards, and the specific circumstances of the engagement, not any inherent hierarchy between the two, determine which is more appropriate.

Question 8

Distinguish random selection, systematic selection, and monetary unit sampling, and explain why monetary unit sampling is particularly well suited to testing for overstatement in a large population of varied-value items.

Accepted Answer

Random selection is a method in which every item in the population has an equal chance of being selected, typically implemented using random number generation to identify which specific items from the population will be examined, with no structured pattern determining the selection beyond genuine randomness. Systematic selection applies a constant sampling interval to the population, such as selecting every fiftieth item, with the starting point within the first interval determined randomly to avoid any systematic bias that might otherwise arise if the interval happened to coincide with some pattern in how the population is ordered. Monetary unit sampling is a value-weighted selection method, in which each individual rupee of value within the population, rather than each discrete item, is treated as a sampling unit with an equal chance of selection; because an item's chance of being selected under this method is proportionate to its rupee value, larger-value items are automatically more likely to be included in the sample than smaller-value items, since a large-value item contains more individual rupees, each of which is a separate chance at selection. Monetary unit sampling is particularly well suited to testing for overstatement in a large population of varied-value items because overstatement risk is naturally concentrated in the larger-value items, an error of a given percentage magnitude produces a larger absolute rupee misstatement on a large item than on a small one, and by weighting selection towards value rather than towards the count of items, the method directs audit effort automatically towards the items where a material overstatement is most likely to be found and where finding it matters most in absolute rupee terms, without requiring the auditor to first stratify the population manually to achieve a similar concentration of effort on higher-value items.

Question 9

State what audit evidence comprises, and give an example of information that is not part of the accounting records but is still relevant audit evidence.

Accepted Answer

Audit evidence is the information used by the auditor in arriving at the conclusions on which the audit opinion is based, and it comprises both information contained in the accounting records underlying the financial statements and other information. The accounting records themselves include the underlying books of account, supporting worksheets, ledgers, journal entries and similar documentation from which the financial statements are derived. Other information, beyond the accounting records, is equally recognised as valid audit evidence and can include minutes of board and committee meetings, confirmations from third parties, industry and economic reports relevant to assessing the reasonableness of management's estimates, and comparable data about competitors used in analytical procedures. An example of information that is not part of the accounting records but is still relevant audit evidence is a set of board meeting minutes documenting a decision to discontinue a product line, which is not an accounting record in itself but is highly relevant evidence bearing on, for instance, whether related inventory should be written down or whether an impairment indicator exists for associated assets, illustrating that the scope of audit evidence extends well beyond the entity's formal accounting books and records alone.

Question 10

An auditor observes a warehouse team performing a year-end inventory count and concludes the counting procedures are being followed correctly throughout the year. Identify the flaw in this conclusion, connecting it to a specific limitation of observation as an audit procedure.

Accepted Answer

The flaw is drawing a conclusion about the entity's counting procedures throughout the year from an observation made at only one point in time, the year-end count itself, when observation as an audit procedure provides evidence only for the specific point in time at which the observation actually occurred, not for any period before or after it. The auditor watching the warehouse team on the day of the count can validly conclude that the counting procedures were being followed correctly on that specific day, but cannot validly extend that conclusion to how, or whether, similar procedures were followed at any other point during the year, since nothing about a single observation reveals what happened on days the auditor was not present to observe. This limitation is compounded by a second, related concern: the observed party may behave differently precisely because they know they are being observed, sometimes described as an observer effect, meaning even the auditor's conclusion about the specific day observed may not perfectly reflect how the same staff would perform the identical task on an ordinary day without an auditor present watching them. Together these two limitations mean observation is valuable evidence for confirming that a procedure, as designed, is capable of being performed correctly and was performed correctly on the specific occasion observed, but it is comparatively weak evidence, on its own, for concluding that the same procedure is reliably and consistently performed throughout an entire period; corroborating this broader conclusion typically requires additional evidence, such as inspection of documented counting records from other dates during the year or inquiry corroborated by other means.

Question 11

Explain what the presentation assertion covers for both classes of transactions and account balances, and give an example of a misstatement that would specifically be a presentation problem rather than an existence, completeness or valuation problem.

Accepted Answer

The presentation assertion, appearing in both the transaction assertions and the balance assertions, concerns whether financial information is appropriately aggregated or disaggregated, clearly described, and whether related disclosures are relevant and understandable in the context of the requirements of the applicable financial reporting framework. It is a distinct assertion from existence, completeness, or valuation because an item can genuinely exist, be complete in its recording, and be correctly valued, and still be misstated purely because of how it has been presented or described in the financial statements, such as being classified under the wrong balance sheet heading or disclosed inadequately. An example of a misstatement that is specifically a presentation problem rather than an existence, completeness or valuation problem: a company correctly records a long-term loan at its correct outstanding amount, the loan genuinely exists in full, the balance is complete and accurately valued, but presents the entire loan as a non-current liability when a substantial instalment is actually due for repayment within the next twelve months and should, under the applicable framework, be reclassified and presented as a current liability. The existence, completeness and valuation of the loan are all correct; the problem lies entirely in how the balance has been presented on the face of the balance sheet, specifically its classification between current and non-current, which affects a reader's assessment of the company's short-term liquidity position even though no underlying figure is factually wrong. This is exactly the kind of assertion-specific problem that requires a presentation-and-disclosure-focused procedure, reviewing the classification and disclosure of balances against the applicable framework's requirements, rather than a procedure aimed at confirming existence or recalculating a value.

Question 12

Distinguish reperformance from recalculation as audit procedures, and give one example of each for testing a payroll process.

Accepted Answer

Recalculation is the audit procedure of checking the mathematical accuracy of documents or records, essentially confirming that arithmetic already performed by the entity has been done correctly, without the auditor independently carrying out the underlying process or control that produced the figures being checked. Reperformance is the auditor's own independent execution of procedures or controls that were originally performed as part of the entity's internal control, going beyond merely checking arithmetic to actually carrying out the substantive process itself and comparing the auditor's own result against what the entity produced. For testing a payroll process, an example of recalculation is checking that an individual employee's gross pay, once the hours worked and the pay rate are taken as given from the entity's own records, has been multiplied correctly and that the resulting statutory deductions have been computed correctly according to the applicable rates, essentially verifying the arithmetic of a calculation the entity has already performed. An example of reperformance is the auditor independently recomputing an employee's entire net pay from first principles, starting from the underlying time records and the applicable pay rate and statutory deduction rules, and comparing the auditor's own independently derived figure against what the entity's payroll system actually paid, which goes further than merely checking the entity's arithmetic and instead reconstructs the whole calculation independently as a genuine, parallel execution of the underlying process, providing stronger evidence than recalculation alone that the entity's payroll process, and not merely its arithmetic, is operating correctly.

Question 13

For a company's trade payables balance, identify which assertion is most likely to be understated risk, and explain why completeness, rather than existence, is typically the assertion of greatest concern for a liability balance in general.

Accepted Answer

For trade payables, the assertion of greatest concern is typically completeness, the risk that not all liabilities that should have been recorded have in fact been recorded, resulting in an understated balance. This reflects a general pattern that applies to liability balances more broadly, and the reasoning is directional: management has an inherent incentive, whether consciously or through the ordinary human tendency to be less diligent about recording unfavourable items, to understate liabilities and thereby present a stronger financial position, since a smaller recorded liability improves reported net assets and can improve reported profit depending on how the corresponding expense is treated. Existence is comparatively less of a concern for a liability balance, because a fictitious liability would be an unusual finding, since there is generally little incentive for management to overstate what the entity owes, in contrast to an asset balance, where the incentive runs the opposite way, and existence risk, the risk of recording an asset that does not actually exist or of overstating an asset's amount, is typically the assertion of greatest concern, since overstated assets similarly present a stronger financial position. This asymmetry, completeness being the primary concern for liabilities and existence being the primary concern for assets, is a recurring and examinable pattern throughout audit practice, and it directly shapes the auditor's choice of procedures: for payables, the auditor places particular emphasis on procedures designed to detect understatement, such as a search for unrecorded liabilities examining payments made shortly after year end to identify obligations that existed at the balance sheet date but were not yet recorded, rather than on procedures aimed at confirming that recorded payables genuinely exist, which would be the natural emphasis for an asset balance instead.

Question 14

Distinguish haphazard selection from random selection, and explain why haphazard selection, though it avoids deliberate bias, does not provide the same statistical basis for evaluating sampling risk that random selection provides.

Accepted Answer

Random selection is a method in which every item in the population has a known, equal, mathematically determined chance of being selected, typically implemented using a random number generator or table to identify the specific items chosen, with the selection process itself following a formal, probability-based procedure. Haphazard selection is an attempt by the auditor to select a sample representative of the population without following any structured technique, deliberately avoiding any conscious bias in which items are chosen, such as avoiding a tendency to pick items that are easiest to locate or that look unusual, but without the formal randomisation process that random selection employs. Because haphazard selection does not follow a defined, probability-based selection mechanism, even though the auditor genuinely intends to avoid bias and may succeed in avoiding conscious bias, there is no mathematical basis for calculating the probability that any given item was selected, and consequently no rigorous statistical basis exists for using the sample results to draw formally quantified conclusions about the whole population, in the way statistical sampling techniques applied to a genuinely random sample allow. Haphazard selection remains an acceptable selection method for non-statistical sampling approaches, where the auditor is relying on judgement rather than formal probability theory to evaluate the results, but it should not be used as the selection method underlying a sampling approach that purports to be statistical, since the absence of a genuinely randomised, probability-defined selection process undermines the validity of the statistical evaluation that a statistical approach depends on.

Question 15

An auditor selects a random sample of 60 sales invoices from a population of 6,000 to test for proper authorisation, and finds 2 invoices lacking the required approval signature. Explain what this result means for the auditor's control risk assessment, the role of sampling risk in interpreting it, and what the auditor should do next rather than simply accepting or rejecting the control based on this raw count alone.

Accepted Answer

Finding 2 unauthorised invoices in a sample of 60 suggests a sample deviation rate of approximately 3.3%, and this raw finding must be interpreted through the lens of sampling risk rather than treated as a precise, direct statement about the true deviation rate across the full population of 6,000 invoices. Because only a sample, not the entire population, was tested, the true rate of unauthorised invoices in the full population could genuinely be higher or lower than the 3.3% observed in the sample, purely as a consequence of which specific 60 items happened to be selected; this is exactly what sampling risk describes, the risk that the auditor's conclusion based on the sample differs from what would be concluded if the entire population were tested. Rather than simply accepting the control as effective because 2 out of 60 sounds like a small number, or rejecting it outright because any deviation was found at all, the auditor should evaluate the sample result formally, comparing the observed deviation rate against a predetermined tolerable deviation rate set before the test was performed, reflecting how much deviation the auditor was prepared to accept and still conclude the control is operating effectively enough to support the planned reliance on it; if using a statistical approach, the auditor would also compute an appropriate upper confidence bound on the true population deviation rate given the sample result, rather than relying on the raw 3.3% observed rate alone, since the upper bound accounts explicitly for sampling risk and gives a more conservative, defensible basis for the conclusion. The auditor should also investigate the nature and cause of the 2 exceptions found, since understanding why they occurred, whether isolated oversights, a systemic weakness at a particular time or location, or a sign of an ineffective control more broadly, is itself important evidence bearing on whether reliance on this control remains appropriate, quite apart from the purely numerical evaluation of the sample result; a projected error rate exceeding the tolerable rate, or exceptions revealing a systemic rather than isolated cause, would lead the auditor to conclude the control is not operating effectively enough to be relied upon, prompting either extended testing or a shift towards a substantive-procedures-only approach for the related assertions instead.

Practice — Audit Evidence and Sampling

Sign in to save attempts to your dashboard and track your accuracy over time

Find a verified tutor

Join a live course

Browse coaching centres

About

Categories

Support

Documentation

Get Study Tips & Updates

Download Our App

Download for iOS

Download for Android

Popular Cities