Information Systems and Data Management
Data foundations
Candidates must separate master data (customers, vendors, chart of accounts), transaction data, and reference data, and know that metadata describes structure and lineage. Data governance assigns owners and policies; data quality dimensions include accuracy, completeness, validity, timeliness, and uniqueness. In relational databases, primary keys uniquely identify rows and foreign keys enforce referential integrity; normalization reduces redundancy, denormalization trades redundancy for query speed. The data dictionary documents fields, types, and valid values. A data warehouse holds structured, modeled, historical data (schema-on-write); a data lake holds raw data (schema-on-read) and needs governance to avoid becoming a swamp.
Controls hierarchy
General IT controls (access management, change management, IT operations) support the reliable functioning of application controls, which are input, processing, and output controls. Input examples: format/field check, validity check (against an authorized list or master file), reasonableness/limit check (against a range), check digit (catches single-digit and transposition errors, not all errors), completeness check, duplicate check. Processing examples: control totals, hash totals (a meaningless sum of a field such as invoice numbers used only for control), record counts, run-to-run balancing. Output examples: distribution logs, review of exception reports.
Traps
Validity versus reasonableness; hash total versus financial total; batch versus online real-time characteristics; and treating data analytics as substantive evidence without testing completeness and accuracy of the source data.
Worked example: To test completeness of recorded sales, reconcile the record count and hash total of invoice numbers from the billing system to the sales journal; to test validity, vouch recorded sales to signed shipping documents.