Classification Drill
Fill every blank before reading the answers.
- Phone number matches approved format but belongs to a different customer → ______
- Mandatory customer email is blank → ______
- ORDER.CUSTOMER_ID has no matching Customer parent → ______
- Feed arrives on schedule, but balances are two months old → ______
- Today’s correct balances arrive three hours after required 8 a.m. deadline → ______
- Same real customer stored twice under separate IDs → ______
- Two systems use same business meaning but conflicting status values → ______
- Daily transaction volume far outside plausible benchmark → ______
- Analyst explores null rates, distributions, patterns, relationships for first time → ______
- Team recalculates same approved DQ score daily and compares with threshold → ______
- Business says every U.S. customer must have ZIP Code → ______
- 97.5% of required ZIP Codes are populated → ______
- At least 99% of required ZIP Codes must be populated → ______
- Severe unresolved issue routed to higher authority because of impact/duration → ______
- Supplier must correct critical defects within four business hours → ______
- Free-form full-name string split into title/given/middle/surname → ______
Tenn.andTNconverted to approved representation → ______- Postal address supplemented with geocode coordinates → ______
- Engine auto-fixes high-confidence records and routes ambiguous cases to Steward → ______
- UI validation and upstream process controls changed so defect is no longer created → ______
Answers
- Accuracy failure (may be valid yet inaccurate).
- Completeness failure.
- Integrity failure.
- Currency failure.
- Timeliness failure.
- Uniqueness/Deduplication failure.
- Consistency failure.
- Reasonableness failure.
- Data profiling.
- Measure and Monitor Data Quality.
- Data Quality rule aligned primarily to Completeness.
- Data Quality metric / measurement result.
- Acceptability threshold / standard.
- Escalation within DQ issue management.
- DQ SLA response/remediation commitment.
- Parsing.
- Standardization / transformation.
- Data enrichment.
- Manually-directed correction.
- Prevention / root-cause remediation.
Mastery rule: a correct label without the deciding reason is not enough. For every miss, state why the correct concept wins and why the nearest alternative loses.