How do I test completeness and accuracy of a system-generated report?

Question

AcadiFi · Accepted Answer

You need evidence that the report captures the right data and reflects the right parameters. Common steps include observing report generation, inspecting filters and date ranges, validating row counts, reconciling totals to a trusted source, and, when needed, understanding the report logic or interface feeding it.

Suppose Harbor Crest Insurance gives internal audit a claims-exception report with 12,480 rows covering Q1. Before using it as the testing population, the auditor should confirm the report parameters, verify the run date, and reconcile the row count to the underlying claims table or another trusted source. If the report pulls only claims above $5,000, that threshold must be confirmed. If the report excludes closed claims by default, that exclusion needs to be intentional. A quick reasonableness test might compare total claim dollars on the report, say $18.6 million, to the claims ledger for the same period.

The most common mistake is assuming a report is reliable because it came from the system. Reports are only as good as their logic, filters, and interfaces. CIA questions often reward the answer that independently verifies parameters and completeness before sampling begins. If the report itself is not trustworthy, the sample built on it will not be trustworthy either. Practice more in our CIA question bank.

How do I test completeness and accuracy of a system-generated report?

Master Part 3 with our CIA Course

Related Questions

Practice Questions