Evidence practice / 3 MIN READ
Write an evidence reliability note before selecting a sample
A fictional change-report case separates a report's appearance from its scope, extraction criteria and ability to support an audit question.
On this page
Before selecting records from a report, establish what population it represents and whether that population answers your question. A sample drawn carefully from an incomplete population does not repair the missing scope.
The European Court of Auditors' guidance on audit evidence treats reliability as dependent on source, type and the circumstances of collection. Use that principle to ask concrete questions; do not turn it into a blanket rule that a system export is always sufficient or a spreadsheet is always unreliable.
Inspect the fictional record
Your practice question is: were production changes during the quarter handled under the applicable approval process? You receive an 80-row file labeled “Q2 completed changes.” The saved query shows a status filter of “closed” and a category filter that excludes emergencies. The file does not explain whether open changes were implemented during the quarter.
| Observation | What you can say | What you cannot yet say |
|---|---|---|
| 80 rows are present | This export contains 80 records | The quarter had exactly 80 production changes |
| Emergency category is excluded | That category is outside this extract | No emergency changes occurred |
| Status is restricted to closed | Other statuses are not represented | Every implemented change was closed |
| Dates appear within Q2 | Displayed dates fall in the period | The date field is the implementation date |
No real employee, customer or system data is needed. Build this exercise with invented records and label it as practice.
Draft a bounded conclusion
“The supplied file is not yet established as the population for the stated question. Its filters exclude emergency changes and non-closed records, and the meaning of the date field is unresolved. I would clarify the extraction criteria and reconcile relevant exclusions before selecting the sample.”
This is not a finding that changes lacked approval. It is a limitation on what you can conclude from the current record. Keep that distinction visible in an interview answer and in the written note.
Specify the next evidence
Request the field definitions, query criteria, extraction time and relevant source scope. Ask how implemented changes with other statuses and emergency changes are represented. Agree the permitted way to reconcile the resulting population to the source; do not request unrestricted production access as a default.
Suppose a revised extract adds 12 emergency records and three implemented records not yet closed, with no overlap among those groups. The practice population is now 95 records, subject to the remaining source checks. The arithmetic reconciles these stated additions; it does not independently verify completeness or prove that approvals were appropriate.
Update your note with the new scope, the reconciliation performed and the questions still open. Preserve the original limitation so the reasoning remains understandable.
Move to selection only when the scope is clear
The sample-selection exercise addresses which records to inspect and the limits of the resulting conclusion. That decision follows this reliability check. Compare evidence-reliability preparation options if you need a larger connected project, or use the free note as your finished practice artifact.
PUT IT INTO PRACTICE
GRC Career Transition Guide
Turn this guidance into a focused role brief, experience inventory and application plan.