Sample Size Calculator

Sampling Considerations for General

Sample size under ISA 530 is half arithmetic, half judgment. The arithmetic is the MUS formula. The judgment is what you set tolerable misstatement to, what confidence factor you pick, and how much expected misstatement you build in. We see two patterns on most files we review. Either the team uses the firm's default confidence factor (2.31 at 90%) without checking whether the assessed risk justifies a higher one, or expected misstatement gets set to zero by default and the sample comes out too small to absorb any real exception. Both fail the same way at inspection: "the rationale for the sample size is not documented."

Sampling focus: General

The MUS formula: n = (Population × Confidence Factor) / (Tolerable Misstatement − Expected Misstatement × Expansion Factor). Confidence factor at 90% is 2.31; at 95% it jumps to 3.00 and your sample roughly doubles. Expected misstatement bites because the formula subtracts it from tolerable, so leaving headroom for known prior-year errors expands the required sample. The expansion factor (1.6 typical) handles the variability around expected misstatement — small input, real impact on n.

Key sampling considerations

Items above the sampling interval are automatically selected in full — they form the top stratum. These items must be tested individually outside the sample.

The confidence factor depends on the assessed risk of material misstatement — higher risk means higher confidence required and larger samples.

Expected misstatement should be based on prior-period results, understanding of the entity, and results of other audit procedures.

Stratification of the population can improve sampling efficiency by reducing variability within each stratum.