Corporate Accounting and Auditing · Audit Sampling, Audit Techniques and Analytical Procedure

Statistical and Non-Statistical Sampling Methods in Audit

Updated 10 October 2026 · Fact-checked

Audit sampling means testing less than 100% of a population and drawing a conclusion about all of it. Statistical sampling uses random selection and probability theory to evaluate results. Non-statistical sampling relies on auditor judgment. Choose the method by population, risk and purpose, then select items by random, systematic, stratified, haphazard or block methods.

Understand Sampling Methods: Statistical and Non-Statistical

Audit sampling means applying audit procedures to less than 100% of the items in a population, so that every item has a chance of being selected. The auditor then draws a conclusion about the whole population. The population is the full set of data from which the sample is drawn, for example all purchase invoices of the year. SA 530 deals with audit sampling.

The approach can be statistical or non-statistical. A statistical sampling approach has two features: the items are chosen randomly, and probability theory is used to evaluate the results, including the sampling risk. If either feature is missing, the approach is non-statistical sampling. Non-statistical sampling depends on the auditor's judgment to decide the size, the selection and the evaluation.

Both approaches can give sufficient appropriate evidence when properly designed. Statistical sampling gives a measurable, objective result and is easier to defend. It needs more training, effort and often software. Non-statistical sampling is cheaper and quicker, but its sampling risk cannot be measured in numbers.

There are several ways to pick the items. Random selection uses random numbers, so each item has an equal chance. Systematic selection takes every nth item after a random start. Stratified selection divides the population into groups with similar features and samples each group. Haphazard selection picks items without a structured technique but without bias. Block (cluster) selection takes a continuous run of items, such as all vouchers of one week.

The real exam skill is matching the method to the situation. Random and systematic suit large, uniform populations. Stratified suits a mixed population with high-value and low-value items. Haphazard suits quick non-statistical tests. Block selection is weak, because a block rarely represents the whole population.

Key rules to remember

Sampling interval (systematic selection): Sampling interval = Population size ÷ Sample size; Pick a random starting point within the first interval, then take every item at that interval. Example: 2,000 items and a sample of 40 give an interval of 50.
Statistical sampling test: Statistical sampling = Random selection + Probability theory to evaluate results; Both conditions are needed. Random selection with judgmental evaluation is non-statistical.
Stratified sample coverage: Items per stratum = Sample size × (Stratum size ÷ Population size), when sampling in proportion; Proportional allocation is one option. Auditors often test all high-value items and sample the rest.
Sampling risk: Sampling risk = risk that the conclusion from the sample differs from the conclusion if the whole population were tested; It arises in both approaches but can be quantified only in statistical sampling.

How to solve Sampling Methods: Statistical and Non-Statistical questions

Use this order for any question asking you to choose, compare or apply a sampling method.

1Identify the population and the audit objective, for example testing existence of debtors or authorisation of purchases.
2Check the population features: size, uniformity, value spread, and whether items are numbered in sequence.
3Decide the approach: statistical if you want a measurable sampling risk and objective evaluation, non-statistical if judgment, cost and time matter more.
4Pick the selection method that fits: random or systematic for uniform populations, stratified for varied values, haphazard for quick judgmental tests, block only when justified.
5Apply the method with numbers where given, such as the sampling interval or items per stratum.
6Evaluate the results, project errors to the population, and state your conclusion.
7Write one line on sampling risk and why the chosen method is suitable.

Quickest way: Two-question shortcut for choosing a method

When to use it: Use in MCQs and short theory questions where you must name the suitable method quickly.

Ask: is probability theory used to evaluate results? If yes, it is statistical. If no, it is non-statistical, even when selection is random.
Ask: what does the population look like? Uniform and numbered points to random or systematic. Mixed values points to stratified. No structure and a quick test points to haphazard. A continuous run points to block.
For systematic problems, compute population ÷ sample first, then list items from the random start.

Common mistakes in Sampling Methods: Statistical and Non-Statistical

Calling any random selection statistical sampling.
Students focus on the word random and ignore the evaluation part.
Fix: Remember both features: random selection and probability-based evaluation. Missing either one makes it non-statistical.
Treating haphazard selection as random selection.
Both sound unplanned, so they look alike.
Fix: Random uses a method such as random numbers, giving every item a known chance. Haphazard has no structure and relies on the auditor avoiding bias, so it is non-statistical.
Choosing systematic sampling without a random start.
Students just start with the first item.
Fix: Choose the starting point randomly within the first interval. Also check that the population has no pattern matching the interval.
Recommending block selection as a reliable basis for conclusions on the whole population.
It is easy to apply, so it looks efficient.
Fix: State that a block rarely represents the population. Use it only with other tests, or when several blocks are taken.
Saying non-statistical sampling is not acceptable or has no sampling risk.
Students assume only statistics gives valid evidence.
Fix: Either approach can give sufficient appropriate evidence. Non-statistical sampling has sampling risk too; it is just not measured numerically.

Worked examples

Example 1

An auditor has 2,400 numbered purchase invoices and wants to test 60 using systematic selection. The random start is invoice 17. List the sampling interval and the first four items selected.

Show the solution

Sampling interval = Population ÷ Sample = 2,400 ÷ 60 = 40.
The random start is 17, which lies within the first interval of 1 to 40.
Second item = 17 + 40 = 57.
Third item = 57 + 40 = 97.
Fourth item = 97 + 40 = 137.

Answer: Interval is 40. The first four invoices are 17, 57, 97 and 137.

Example 2

Debtors of Sundaram Traders Ltd. total 500 accounts. 20 accounts, each above ₹5,00,000, are 70% of the balance. 130 accounts are between ₹1,00,000 and ₹5,00,000. The remaining 350 are small. Suggest a stratified approach, testing all large accounts, 10% of the middle stratum and 2% of the small stratum.

Show the solution

Form three strata: large (20 accounts), medium (130 accounts), small (350 accounts).
Large stratum: these cover most of the value and carry high risk, so test all 20 accounts.
Medium stratum: 10% of 130 = 13 accounts selected randomly.
Small stratum: 2% of 350 = 7 accounts selected randomly.
Total sample = 20 + 13 + 7 = 40 accounts.
Reason: stratification lets the auditor focus effort on high-value items and reduces the variability within each group.

Answer: Test all 20 large accounts, 13 medium accounts and 7 small accounts, a total of 40 accounts out of 500.

Exam tips

For difference questions, set out a two-column comparison: selection method, evaluation method, sampling risk measurement, cost and skill needed.
In numerical questions, always show the sampling interval formula and the random start before listing items.
For suitability questions, link the method to the population: stratified for varied values, systematic for numbered uniform items, block only with caution.
Write that SA 530 governs audit sampling, and mention sampling and non-sampling risk in one line to earn extra marks.
In MCQs, check whether the question mentions probability-based evaluation; that decides statistical versus non-statistical.

Practice questions from Audit Sampling, Audit Techniques and Analytical Procedure

Sampling Methods: Statistical and Non-Statistical: frequently asked questions

What is the main difference between statistical and non-statistical sampling in audit?

Statistical sampling uses random selection and probability theory to evaluate results, so sampling risk can be measured. Non-statistical sampling relies on the auditor's judgment for selection and evaluation. Both can give valid audit evidence.

What is the difference between random and systematic sampling in auditing?

Random sampling picks items using random numbers, so each item has an equal chance. Systematic sampling picks every nth item after a random start, where n is the population divided by the sample size. Systematic is quicker but can mislead if the population has a repeating pattern.

When is stratified sampling used in audit?

It is used when the population has items of very different values or risk, such as debtors or inventory. The auditor divides the items into groups and samples each group, often testing all large items. This improves efficiency and coverage.

Is haphazard sampling statistical?

No. Haphazard selection has no structured technique, so probability theory cannot be applied. The auditor must still try to avoid bias, for example by not choosing only easy or unusual items.