What it is
Attribute sampling answers one question for a control: how often does it fail? From the tolerable deviation rate, the expected rate and the risk of overreliance you get a sample size; from the deviations found you get an upper deviation limit to compare with the tolerable rate.
This workbook holds the sample size tables for 90% and 95% confidence and tolerable rates from 2% to 20%, calculated from the exact binomial distribution. It picks the sample size for the expected rate, checks it with BINOMDIST, builds a systematic selection with a random start, records results for up to three attributes and computes the exact upper deviation limit.
When to use it
- Sizing and evaluating a test of controls under AU-C 330 and AU-C 530.
- Compliance testing where each item either meets the attribute or does not.
- Showing the basis for a sample size in the file.
What’s inside
- Sample size lookup by confidence, tolerable rate, expected rate and deviations accepted
- Self-check of the chosen size against the binomial distribution
- Selection of up to 500 items with a random start and result entry for three attributes
- Upper deviation limit by attribute and a conclusion against the tolerable rate
Standard cited
This template follows AU-C 530 / AU-C 330. It is written in our own words, except for phrases the standard requires the document to contain, with [bracketed placeholders] for client details, amounts in USD and dates as MM/DD/YYYY. Check it against the current standard and your firm’s methodology before use. See AU-C 530 explained, AU-C 330 explained.
Questions
Do the sizes match the AICPA tables?
Each size is the smallest sample for which the exact binomial chance of finding no more than the accepted number of deviations, if the true rate equals the tolerable rate, is no more than the risk of overreliance. At 95% confidence and a 5% tolerable rate the sizes for 0 to 5 accepted deviations are 59, 93, 124, 153, 181 and 208; at 90% confidence and 5% tolerable with no deviations it is 45. The tables are calculated, not copied, and the workbook rechecks the chosen size with BINOMDIST. A firm table built on a different approximation can differ by an item or two, so use the table your methodology prescribes.
What is the upper deviation limit?
The highest population deviation rate that is still plausible given the sample, at your confidence level. With no deviations in 59 items at 95% confidence it is just under 5%. If it exceeds the tolerable rate, the sample does not support the planned reliance on the control.
Does the workbook handle small populations?
It includes the optional adjustment for a sample above 10% of the population. For very small populations, test every item. Many firms set minimum samples by control frequency; apply the larger of that and the statistical size.
Related templates
- SOX Control Test Workpaper (TOD / TOE)
- Monetary Unit Sampling Workpaper (MUS)
- Audit Risk Assessment Worksheet (SAS 145)
- Payroll Testing Workpaper
Last updated 11/05/2026. All free templates.