toolbox
Home / Legal, Safety & Compliance / Patent Maintenance Audit Sample Size Estimator

Patent Maintenance Audit Sample Size Estimator

This estimator calculates a statistical sample size for auditing a finite population of patent-maintenance records. It helps compliance or IP operations teams plan spot checks when they want a sample capable of estimating an exception rate within a chosen margin of error and confidence level.

The result is a statistical planning aid rather than a mandatory legal sample. Higher confidence, a smaller margin of error, and an expected exception rate near 50% generally increase the sample size; a finite-population correction reduces the sample when the total record population is limited.

Patent maintenance audit sampling assumptions

%
%
Result
—
Recommended statistical sample size
Uncorrected sample—
Population sampled—
Expected exceptions in sample—
Confidence level—

1. Enter the population
Use the total number of patent-maintenance records eligible for the audit period.

2. Choose confidence
Select the statistical confidence level used for the estimate.

3. Set precision
Enter the maximum margin of error you are willing to accept for the estimated exception rate.

4. Estimate the exception rate
Use a reasonable prior expectation. If there is no basis for an estimate, 50% is conservative for sample-size planning.

5. Use the rounded sample
The calculator rounds the statistical result up to a whole record and limits it to the full population.

Formula:

n₀ = z² × p × (1 − p) ÷ e²; n = n₀ ÷ [1 + (n₀ − 1) ÷ N]

z = z-score for the selected confidence level
p = expected exception proportion as a decimal
e = margin of error as a decimal
N = finite population size
n = finite-population sample size

Assumptions: This is a simple random-sampling model for estimating a proportion. Risk-based, stratified, discovery, or regulatory sampling plans may require a different method.

What the result means

The sample size is the number of records suggested by the selected statistical assumptions. It does not establish that the audit design meets a regulator, court, or internal policy requirement.

Consider professional statistical or legal guidance when the sample supports a consequential compliance conclusion.

Given

  • Population N = 1,000 records
  • 95% confidence, z = 1.96
  • Margin of error e = 5% = 0.05
  • Expected exception rate p = 10% = 0.10

Calculation
n₀ = 1.96² × 0.10 × 0.90 ÷ 0.05² ≈ 138.30. Finite correction: n = 138.30 ÷ [1 + (137.30 ÷ 1,000)] ≈ 121.60. Round up to 122.

Result
Recommended sample: 122 records.

Under these assumptions, reviewing 122 randomly selected records provides the chosen statistical precision for estimating the exception proportion.

Why can a 50% expected exception rate produce a larger sample?

For a proportion estimate, p × (1 − p) is largest at 50%. When there is no reliable prior estimate, 50% therefore gives a conservative sample size.

Should I use 99% confidence for every audit?

Not automatically. Higher confidence increases sample size, so select a level appropriate to the audit objective, risk, and governing methodology.

Can I use this for a targeted high-risk sample?

The formula assumes simple random sampling. A targeted or risk-based sample can be useful operationally but does not have the same statistical interpretation.

What if the calculated sample is larger than the population?

The tool caps the recommendation at the full population because you cannot sample more unique records than exist.

Does this sample size prove patent-maintenance compliance?

No. Sample size addresses statistical coverage only; compliance conclusions also depend on the review criteria, evidence quality, and applicable rules.