---
title: "Plan audit sampling and record conclusions"
description: "Define the population, record sampled evidence and preserve changes in sampling rationale."
url: https://helpcenter.complytrain.net/en/articles/91-plan-audit-sampling-and-record-conclusions
collection: "Audits and management review"
updated: 2026-09-22
language: en
---

# Plan audit sampling and record conclusions

Define the population, record sampled evidence and preserve changes in sampling rationale.

Define the population, record sampled evidence and preserve changes in sampling rationale.

## Start a sampling plan

Product location: `/qms/audits`. Select Audits, open the engagement and choose Sampling. The tab and selected record are not encoded in this URL.

The QMS audit-management permission is required to view or change sampling plans. Open the audit engagement and **Sampling**, then **New sampling plan**. Define the population you intend to examine before adding individual samples. For example, identify the period, locations and type of records that the population includes.

<table style="min-width: 50px;"><colgroup><col style="min-width: 25px;"><col style="min-width: 25px;"></colgroup><tbody><tr><th colspan="1" rowspan="1"><p>Field</p></th><th colspan="1" rowspan="1"><p>What to enter</p></th></tr><tr><td colspan="1" rowspan="1"><p>Plan name</p></td><td colspan="1" rowspan="1"><p>A required name, up to 300 characters, that distinguishes the purpose of the sample.</p></td></tr><tr><td colspan="1" rowspan="1"><p>Population</p></td><td colspan="1" rowspan="1"><p>Required description of what is being sampled.</p></td></tr><tr><td colspan="1" rowspan="1"><p>Population size</p></td><td colspan="1" rowspan="1"><p>Optional whole-number count, zero or greater.</p></td></tr><tr><td colspan="1" rowspan="1"><p>Method</p></td><td colspan="1" rowspan="1"><p>Random, Systematic or Judgmental. Random is initially selected.</p></td></tr><tr><td colspan="1" rowspan="1"><p>Sample size</p></td><td colspan="1" rowspan="1"><p>Required positive whole number.</p></td></tr><tr><td colspan="1" rowspan="1"><p>Rationale</p></td><td colspan="1" rowspan="1"><p>Why the method and sample size fit this engagement.</p></td></tr><tr><td colspan="1" rowspan="1"><p>Status</p></td><td colspan="1" rowspan="1"><p>Draft is the initial state. Select Active when using the plan; saving an Active plan is not evidence that sampling has occurred.</p></td></tr></tbody></table>

Choosing a method records your intended approach. This form does not itself draw a random sample or establish that a sample size is sufficient for an audit. Record how the selection was actually made and the limitations of the resulting evidence.

All sizes must be whole numbers no greater than 2,147,483,647. The UI requires a positive sample size. An unknown population can be left blank; entering zero means a recorded size of zero. Check that the intended sample and subgroup targets make sense against the actual population rather than relying on numeric validation alone.

Select **Save**, reopen the engagement and confirm the plan name, method, status, population and rationale. **Edit** changes the current editable plan. Use **New version** when the previous plan needs to remain as the record of an earlier sampling round.

![This illustrative draft uses a judgmental sample of five records from a population of twenty. The rationale must explain the organisation’s actual choice.](https://staticfiles.gleap.io/p6950dfbf97796f2087a9f022/mQ2imYuzqWlyPlbv2OczUtUcamcKE5Q7mcMVwOSggDmevotNQoeGZ1tGKD4rW38favkWnBRg4RC.png "Define a sample and its rationale")![The run-059 plan remains Draft after reopening the engagement. Evidence is 0% and no conclusion has been recorded; the earlier example is a separate plan.](https://staticfiles.gleap.io/p6950dfbf97796f2087a9f022/wBqQHjKObfkO9nT077zymtmbdWMlgymflOKzeUl1BvZZWhOXtCw7tj0uTvyjwojZ1OFCbINcKzN.png "Reopen the saved sampling plan")![The active run-115 plan samples three of twelve synthetic records using a stated judgmental rationale. It is separate from the earlier Draft example.](https://staticfiles.gleap.io/p6950dfbf97796f2087a9f022/LcmUySS6MXpLaYgO0pqTVRnbjFdfuD3KSxQhP31o6xjS2OXTdXc0NUQp5pkIuIT8dzx408xG9me.png "Define a fresh sampling round")

## Divide the population into strata

Product location: `/qms/audits`. Select Audits, open the engagement and choose Sampling. The tab and selected record are not encoded in this URL.

Expand the plan and select **Add stratum**. A stratum is a defined subset of the population, such as one site, process or period. Enter its required **Name** (up to 300 characters), a **Definition** and an optional nonnegative **Target size**. The form also offers conclusion fields; leave these not concluded until the subgroup has been reviewed. Select **Save** and inspect the resulting stratum.

Use definitions that let another reviewer understand which records belong in each subset. The target count and the actual evidence collected are separate values. A target of zero or an omitted target does not establish a completed review.

![The stratum names engineering change records, describes its boundary and sets a target of three. The conclusion is still unset.](https://staticfiles.gleap.io/p6950dfbf97796f2087a9f022/dcMC524UkkhwwrVA9x6mJGO9GvxSq36TDmFzSFNHWWiymLfSDplkfN1ZlIgeANCexZ59IWtx2jL.png "Define the subgroup and target")

## Record sampled items

Product location: `/qms/audits`. Select Audits, open the engagement and choose Sampling. The tab and selected record are not encoded in this URL.

Within the relevant stratum, select **Add item**. Supply an **Item reference**, select **Linked evidence**, or provide both. The evidence selector is drawn from the engagement's evidence.

Enter **Notes** that explain what was inspected. Choose **Conforming** or **Nonconforming** when evaluated, or leave the item **Not evaluated** until the check is done. A linked evidence record does not automatically supply an auditor's evaluation.

The item reference is limited to 300 characters. Select **Save**, then check the row under the intended stratum. Its edit icon changes the reference, link, result or notes; it does not change the underlying evidence file.

The UI prevents saving with neither a reference nor an evidence link. If expected evidence is absent, confirm the engagement and its evidence collection; an empty selector can also result from a failed evidence load.

![EC-014 has a Nonconforming result, the three-record evidence file and an explanatory note. Other sampled items need their own result.](https://staticfiles.gleap.io/p6950dfbf97796f2087a9f022/oP4tR7nBhzgsjFhvH5FTrLgqbNLjBLJUL0ZbB3saP7BuSKgaUKo6FBRivXj0mBWothmFCrvzluN.png "Link each inspected item to evidence")

## Read the collection indicators

Product location: `/qms/audits`. Select Audits, open the engagement and choose Sampling. The tab and selected record are not encoded in this URL.

The **Evidence** percentage counts sampled items with an evidence link divided by the plan's intended sample size, rounded and capped at 100%. A reference-only item does not count as collected evidence for this indicator. Linking one evidence file to three distinct sampled items can count as three linked items; the indicator is not a count of unique files.

The **Concluded** percentage counts strata with a recorded conclusion result divided by all strata. Conforming, Nonconforming and Inconclusive all count as a recorded conclusion. No strata produces an unavailable conclusion percentage rather than a completed review.

The demonstration links three items to a file containing all three synthetic entries. Evidence reaches 100%, while Concluded remains 0% until the auditor saves a stratum conclusion. One sampled item is nonconforming: 100% collection does not mean 100% conformity.

![Three evidence-linked items produce Evidence 100%, while Concluded remains 0%. One item is nonconforming despite complete collection.](https://staticfiles.gleap.io/p6950dfbf97796f2087a9f022/lalgxw2YKS9JZYoZ9J9wsmA6TbTreJD4CluNIBsbxh9467VJE0fUDFKrZMwgR1ZgOg9LNAviM3W.png "Separate collection from conclusion")

## Conclude on each stratum

Product location: `/qms/audits`. Select Audits, open the engagement and choose Sampling. The tab and selected record are not encoded in this URL.

Use the stratum’s edit icon to record **Conclusion result** and **Conclusion notes**, then select **Save**. The result can be **Conforming**, **Nonconforming** or **Inconclusive**; an unset result remains **Not concluded**.

Read the plan's **Evidence** and **Concluded** indicators separately. Collected samples and recorded conclusions measure different work. They do not establish that an entire standard conforms. Record a finding through the engagement workflow when the evidence warrants one, and trace it back to the sampled item.

Reopen the engagement and expand the plan to confirm the saved conclusion and supporting item results. In the example, one identifier mismatch leads to a Nonconforming stratum conclusion. Both collection and conclusion indicators reach 100%, while the finding still needs its separate response and verification.

![The auditor records a Nonconforming conclusion and explains the one mismatch before saving.](https://staticfiles.gleap.io/p6950dfbf97796f2087a9f022/IHgyeNl9JEHUo6yT8Bnk0XrBMEpCY5FJh7zwczfNX8NgSazsiXQX9mHKJMkrPWyRqNLJxxfaibI.png "Record the subgroup’s conclusion")![The Nonconforming conclusion and note persist. Both indicators show 100% because collection and conclusion recording are complete, not because all items conform.](https://staticfiles.gleap.io/p6950dfbf97796f2087a9f022/FxevGERwXbHObYx36JKrGHhAkz7tBtS9y5UIqdt1LeMptQjAxzmsVoqfUnQ2rHagEL3NHAOHw6M.png "Reopen the reviewed sampling round")

## Revise the rationale without losing the earlier plan

Product location: `/qms/audits`. Select Audits, open the engagement and choose Sampling. The tab and selected record are not encoded in this URL.

Use **New version** when the existing plan's rationale needs to remain as the earlier audit record. The confirmation creates a new editable version and marks the current plan **Superseded**. The new plan starts **Active** with an incremented version number and a link to the earlier version. The name, population, method, sizes and rationale are copied. Stratum names, definitions and targets are copied, but sampled items and stratum conclusions are cleared for the new round.

A superseded plan is read-only; inspect the version and status before continuing. The previous round retains its items and conclusions when expanded. Its live collection percentages are no longer shown by the active-plan coverage view. The new version starts with empty collection and pending conclusions. Version the current plan, not a superseded one; a concurrent versioning attempt can be rejected when another user has already superseded it.

![The confirmation explains that the current plan becomes Superseded and a new editable version is created.](https://staticfiles.gleap.io/p6950dfbf97796f2087a9f022/8uuFssG0LpybC8aAqJXv9FtAeJJgTJO5M3eM5VFo3ERawn4QJkdIggt5EC9E4kxBkQHcOctBJyY.png "Review the versioning decision")![Version 1 is Superseded and read-only. Version 2 is Active with empty collection and pending conclusions, while retaining the plan’s population and rationale.](https://staticfiles.gleap.io/p6950dfbf97796f2087a9f022/PuZJIZIjWHgV2MCYVB2v2K6GXq2tdv7GOrMl7ppgfIcVurZQ0W1EiWsIDZID6JDmbSAa38caEUs.png "Keep the earlier round and start the next")

## Correct or remove sampling records

Product location: `/qms/audits`. Select Audits, open the engagement and choose Sampling. The tab and selected record are not encoded in this URL.

**Delete** permanently removes a plan and its strata and items. Deleting a stratum removes its sampled items. These actions differ from superseding a plan and their confirmations say they cannot be undone.

After saving or versioning, reopen the plan and check its population, sample size, strata and conclusions. If a save fails, correct the stated field or operation error before repeating the action.

A sampling item can only link evidence from its own engagement. If a save fails because the evidence belongs elsewhere, locate or collect the correct engagement evidence. Do not change the audit’s population simply to use an available file. If a plan became Superseded while editing, reopen the current version and review what was carried forward before making a change.
