Audit Sampling Interview Questions & Answers

Q: How do you evaluate the results of a control test sample?

Count the deviations (control failures) in the sample, compute the sample deviation rate, and for statistical sampling determine the upper deviation rate (adding an allowance for sampling risk). If the upper deviation rate is at or below the tolerable rate, the auditor may rely on the control; if it exceeds, control risk is higher and substantive procedures increase.

Q: What do you do when a sample reveals a misstatement or deviation?

Investigate the nature and cause of each exception (error vs fraud, isolated vs systematic). Consider whether it's an anomaly (rare, non-representative) which requires strong justification and extra work; otherwise project it. Reassess whether related risks and controls are affected, and extend procedures if the projected misstatement approaches or exceeds tolerable misstatement.

Q: What is a representative sample?

A representative sample is one whose characteristics reflect those of the population, so conclusions drawn from it validly apply to the whole. Achieving representativeness requires appropriate selection (giving items a chance of selection) and adequate size. A biased or too-small sample can mislead the auditor about the population.

Q: What is attribute sampling versus variables sampling?

Attribute sampling tests the rate of occurrence of a characteristic—used in tests of controls to estimate the deviation rate (a yes/no attribute like 'was it approved?'). Variables sampling estimates a numerical amount—used in substantive testing to estimate monetary misstatement. Attribute answers 'how often?'; variables answers 'how much?'.

Q: When is 100% testing or specific-item selection used instead of sampling?

Test 100% when the population is small, each item is high value, or a significant risk warrants it and data analytics make full testing feasible. Specific-item selection targets particular items (all items over a threshold, unusual or suspicious items, or items for information)—but testing selected items isn't sampling because conclusions don't extend to the rest of the population.

Q: How does the confidence level relate to sampling?

The confidence level is the complement of sampling risk—e.g., 95% confidence means a 5% sampling risk of a wrong conclusion. Higher required confidence (lower acceptable sampling risk) increases the sample size. The auditor sets confidence based on how much assurance the sample must provide given other evidence and assessed risk.

Q: What are the limitations and risks of relying on sampling?

Sampling inherently accepts sampling risk—the sample may not reflect the population, missing errors (over-reliance/incorrect acceptance) or wrongly rejecting good populations. Poorly designed samples, unrepresentative selection, small sizes, or ignoring the population's nature undermine conclusions. Sampling also can't reliably detect rare but material items unless value-weighted or specifically targeted.

Q: How does reliance on internal controls affect substantive sample sizes?

If tests of controls support a lower control-risk assessment, the auditor can accept higher detection risk and reduce the extent (sample size) of substantive procedures. Conversely, weak or untested controls mean higher control risk, lower acceptable detection risk, and larger substantive samples. Control reliance and substantive extent are inversely related.

How do you evaluate the results of a control test sample?

Intermediate

Count the deviations (control failures) in the sample, compute the sample deviation rate, and for statistical sampling determine the upper deviation rate (adding an allowance for sampling risk). If the upper deviation rate is at or below the tolerable rate, the auditor may rely on the control; if it exceeds, control risk is higher and substantive procedures increase.

Sample 60, deviations 2 -> 3.3% sample rate;
if upper deviation rate <= tolerable rate -> rely on control.

Real-world example Finding 2 exceptions in 60 items, the auditor computes the upper deviation rate and decides whether reliance is still justified.

Common follow-ups: What is the upper deviation rate? | What if deviations exceed tolerable?

Internal Controls Evaluation Audit Risk & Materiality Audit Sampling

What do you do when a sample reveals a misstatement or deviation?

Advanced

Investigate the nature and cause of each exception (error vs fraud, isolated vs systematic). Consider whether it's an anomaly (rare, non-representative) which requires strong justification and extra work; otherwise project it. Reassess whether related risks and controls are affected, and extend procedures if the projected misstatement approaches or exceeds tolerable misstatement.

Real-world example A deviation traced to one new clerk's training gap is investigated to confirm it isn't systematic before concluding.

Common follow-ups: When can a misstatement be treated as an anomaly? | Why investigate the cause of exceptions?

Audit Risk & Materiality Fraud & Error Responsibilities Audit Sampling

What is a representative sample?

Beginner

A representative sample is one whose characteristics reflect those of the population, so conclusions drawn from it validly apply to the whole. Achieving representativeness requires appropriate selection (giving items a chance of selection) and adequate size. A biased or too-small sample can mislead the auditor about the population.

Real-world example A properly randomized sample of invoices mirrors the population's error rate, supporting a valid conclusion.

Common follow-ups: What threatens representativeness? | How does selection method affect it?

Audit Evidence & Procedures Audit Risk & Materiality Audit Sampling

What is attribute sampling versus variables sampling?

Advanced

Attribute sampling tests the rate of occurrence of a characteristic—used in tests of controls to estimate the deviation rate (a yes/no attribute like 'was it approved?'). Variables sampling estimates a numerical amount—used in substantive testing to estimate monetary misstatement. Attribute answers 'how often?'; variables answers 'how much?'.

Real-world example Attribute sampling measures how often approvals were missing; variables sampling estimates the dollar misstatement in balances.

Common follow-ups: Which is used for controls vs substantive? | What does each estimate?

Internal Controls Evaluation Audit Evidence & Procedures Audit Sampling

When is 100% testing or specific-item selection used instead of sampling?

Intermediate

Test 100% when the population is small, each item is high value, or a significant risk warrants it and data analytics make full testing feasible. Specific-item selection targets particular items (all items over a threshold, unusual or suspicious items, or items for information)—but testing selected items isn't sampling because conclusions don't extend to the rest of the population.

Real-world example The auditor tests all ten material journal entries individually rather than sampling, given their size and risk.

Common follow-ups: Does specific-item testing let you conclude on the whole population? | When is 100% testing efficient?

Audit Evidence & Procedures Audit Risk & Materiality Audit Sampling

How does the confidence level relate to sampling?

Beginner

The confidence level is the complement of sampling risk—e.g., 95% confidence means a 5% sampling risk of a wrong conclusion. Higher required confidence (lower acceptable sampling risk) increases the sample size. The auditor sets confidence based on how much assurance the sample must provide given other evidence and assessed risk.

Real-world example Requiring 95% rather than 90% confidence for a high-risk area increases the needed sample size.

Common follow-ups: How does confidence relate to sample size? | What is the complement of confidence?

Audit Risk & Materiality Audit Evidence & Procedures Audit Sampling

What are the limitations and risks of relying on sampling?

Advanced

Sampling inherently accepts sampling risk—the sample may not reflect the population, missing errors (over-reliance/incorrect acceptance) or wrongly rejecting good populations. Poorly designed samples, unrepresentative selection, small sizes, or ignoring the population's nature undermine conclusions. Sampling also can't reliably detect rare but material items unless value-weighted or specifically targeted.

Real-world example A rare but material fraudulent transaction might not appear in a random sample, so it's targeted via specific selection or analytics instead.

Common follow-ups: How can rare material items evade sampling? | What design flaws undermine sampling?

Audit Risk & Materiality Fraud & Error Responsibilities Audit Sampling

How does reliance on internal controls affect substantive sample sizes?

Intermediate

If tests of controls support a lower control-risk assessment, the auditor can accept higher detection risk and reduce the extent (sample size) of substantive procedures. Conversely, weak or untested controls mean higher control risk, lower acceptable detection risk, and larger substantive samples. Control reliance and substantive extent are inversely related.

Real-world example Effective, tested purchase controls let the auditor reduce the substantive sample of purchase transactions.

Common follow-ups: How does control reliance change detection risk? | What happens to samples if controls are weak?

Internal Controls Evaluation Audit Risk & Materiality Audit Sampling