## SA 530: Statistical vs. Non-Statistical Sampling
### Two Approaches to Audit Sampling
Statistical Sampling:
- Uses random selection of sampling units
- Applies probability theory to evaluate results
- Sampling risk can be measured and quantified
- Results can be projected onto the population with a known confidence level
Non-Statistical Sampling:
- Sample selection and evaluation based on auditor's professional judgment
- No mathematical framework for projecting results
- Sampling risk exists but cannot be measured
### Why Statistical Sampling Is Preferred for Large, Homogeneous Populations
1. Mathematical rigour: Sample sizes are determined using laws of probability, not personal judgment, making the process more defensible.
2. Eliminates personal bias: No individual auditor's preferences or preconceptions influence which items are selected.
3. Projectable results: Since selection is random and probability-based, sample conclusions can be mathematically extended to the whole population.
4. Appropriate for large volumes: In large organisations with thousands of similar transactions (payroll, petty cash, invoices, trade receivables), statistical sampling ensures fair coverage.
5. Compliance and substantive testing: Statistical sampling has wide application in both tests of controls (compliance testing) and tests of details.
### Common Applications
- Trade receivables confirmation
- Payroll checking
- Vouching of purchase invoices
- Petty cash voucher verification
### Important Note
Both approaches, if properly applied, can provide sufficient appropriate evidence. SA 530 does not mandate statistical sampling — but it is generally more defensible and scientifically sound for large populations.