CDMP DMF Practice Test 19
Question 180
After addressing this problem, what evidence would justify closure? A scientific dataset has fine precision but lacks metadata required for the regulator's audit.
- Every accepted line references a valid invoice or has an owned exception.
- The baseline includes measured distributions, anomalies and reliability of sampling.
- Regulatory users can reproduce lineage and assess permitted use of each observation.
- The proportion of applications with credible required income rises under a defined rule.
Explanation
Correct Answer
C:It tests the action: Evaluate quality against the specific consumer's required provenance and interpretability.
Incorrect Answers
A:Every accepted line references a valid invoice or has an owned exception. This would verify a different control (An invoice line appears without an invoice header after a partial load.), not the one in the question. The required evidence is: Regulatory users can reproduce lineage and assess permitted use of each observation.
B:The baseline includes measured distributions, anomalies and reliability of sampling. This would verify a different control (A data team wants to set quality targets before knowing null, duplicate and value distributions.), not the one in the question. The required evidence is: Regulatory users can reproduce lineage and assess permitted use of each observation.
D:The proportion of applications with credible required income rises under a defined rule. This would verify a different control (Loan applications often omit annual income even though policy requires it for underwriting.), not the one in the question. The required evidence is: Regulatory users can reproduce lineage and assess permitted use of each observation.
Question 181
Consider this data-management scenario: A CRM has duplicate contact records with different punctuation in their names. Which underlying principle MOST directly explains the preferred response?
- Uniqueness depends on the real-world entity being represented, not only string equality.
- Validity measures conformance to allowed formats, domains and business rules.
- Data-quality issue processes should prioritize consequential defects and accountable response.
- High performance on one data-quality dimension does not establish overall fitness.
Explanation
Correct Answer
A:Uniqueness depends on the real-world entity being represented, not only string equality. This concept explains the reason for the proposed control. Appropriate intervention: Define business uniqueness and controlled entity matching rather than exact-row equality.
Incorrect Answers
B:This is a defensible principle in another context (A product status field contains values archived and disabled although only active and retired are approved.), not the most relevant rationale here. For the stated issue: Uniqueness depends on the real-world entity being represented, not only string equality.
C:Data-quality issue processes should prioritize consequential defects and accountable response. This is a defensible principle in another context (High-impact reconciliation defects are buried among hundreds of cosmetic warnings.), not the most relevant rationale here. For the stated issue: Uniqueness depends on the real-world entity being represented, not only string equality.
D:This is a defensible principle in another context (A postal address field is 100 percent complete but several entries are no longer current.), not the most relevant rationale here. For the stated issue: Uniqueness depends on the real-world entity being represented, not only string equality.
Question 182
A data analyst must confirm the cause of this issue: A company keeps only the latest backup copy, leaving no way to recover before ransomware encryption. Which check provides the strongest diagnostic evidence?
- Inspect slow query logs, wait statistics, resource metrics and alert history.
- Inspect backup immutability, retention history and recovery point selection.
- Review disk growth curves, peak utilization and maintenance approvals.
- Review patch matrix, test plan, recovery procedure and production results.
Explanation
Correct Answer
B:Inspect backup immutability, retention history and recovery point selection. This test checks the relevant business and technical failure rather than assuming its cause. Recovery needs retained trustworthy recovery points as well as available replicas. Expected control result: A pre-incident restore point survives attack and recovers clean data.
Incorrect Answers
A:Inspect slow query logs, wait statistics, resource metrics and alert history. This check would investigate another issue (A database service is slow but monitoring records only whether its process is up.), not the immediate case. The direct diagnostic is: Inspect backup immutability, retention history and recovery point selection.
C:Review disk growth curves, peak utilization and maintenance approvals. This check would investigate another issue (A database approaches disk exhaustion during a seasonal reporting surge.), not the immediate case. The direct diagnostic is: Inspect backup immutability, retention history and recovery point selection.
D:Review patch matrix, test plan, recovery procedure and production results. This check would investigate another issue (A security patch must be applied to a data store with a strict availability commitment.), not the immediate case. The direct diagnostic is: Inspect backup immutability, retention history and recovery point selection.
Question 183
A proposed shortcut is offered for this situation: A data team wants to set quality targets before knowing null, duplicate and value distributions. Which choice is the CLEAREST mistake?
- Check relevant business definitions before joining independently sourced data.
- Invent target percentages without examining actual data distributions.
- Inspect lineage and source context before changing a derived report.
- Validate critical changes against representative positive and negative examples.
Explanation
Correct Answer
B:Invent target percentages without examining actual data distributions. Data profiling provides empirical evidence before quality thresholds are selected. A sound response instead: Run repeatable data profiling with representative population and documented limitations.
Incorrect Answers
A:Check relevant business definitions before joining independently sourced data. Semantic agreements prevent incorrect cross-system correlations. The case-specific response is: Run repeatable data profiling with representative population and documented limitations.
C:Inspect lineage and source context before changing a derived report. Provenance analysis distinguishes source errors from transformations. The case-specific response is: Run repeatable data profiling with representative population and documented limitations.
D:Validate critical changes against representative positive and negative examples. Controlled tests provide more reliable proof than untested assumptions. The case-specific response is: Run repeatable data profiling with representative population and documented limitations.
Question 184
In reviewing this data risk, select the LEAST defensible practice: A registration field is blank, but downstream calculations treat unknown age as zero.
- Check relevant business definitions before joining independently sourced data.
- Define the intended data population before interpreting quality percentages.
- Monitor control outcomes after deployment rather than relying only on design intent.
- Replace every missing age with zero without documenting assumptions.
Explanation
Correct Answer
D:Replace every missing age with zero without documenting assumptions. Null can denote absent or unknown information and is not automatically a numeric zero. A sound response instead: Represent missingness according to explicit semantic rules and validate downstream behavior.
Incorrect Answers
A:Check relevant business definitions before joining independently sourced data. Semantic agreements prevent incorrect cross-system correlations. The case-specific response is: Represent missingness according to explicit semantic rules and validate downstream behavior.
B:Define the intended data population before interpreting quality percentages. An undisclosed denominator can make strong scores misleading. The case-specific response is: Represent missingness according to explicit semantic rules and validate downstream behavior.
C:Monitor control outcomes after deployment rather than relying only on design intent. Operations need evidence that a planned control actually works. The case-specific response is: Represent missingness according to explicit semantic rules and validate downstream behavior.
Question 185
The data management team encounters this situation: A published regulatory dashboard cannot explain why last month's total changed after a source correction. Which intervention BEST addresses the actual problem?
- Implement incremental extraction with reliable change checkpoints and periodic full reconciliation.
- Preserve run versions, source snapshots and transformation lineage with approved restatement rules.
- Use Type 2 dimension rows with effective dates, surrogate keys and validity checks.
- Provide certified datasets, semantic layer guidance and role-based BI access.
Explanation
Correct Answer
B:Preserve run versions, source snapshots and transformation lineage with approved restatement rules. Trusted BI requires reproducible metrics and transparent revision histories. Evidence to obtain: Auditors reproduce the original total and explain the changed figure from documented inputs.
Incorrect Answers
A:This response fits a separate situation (A nightly warehouse load rereads the entire history and repeatedly misses its completion window.), but does not resolve the stated case. Implement incremental extraction with reliable change checkpoints and periodic full reconciliation.
C:Use Type 2 dimension rows with effective dates, surrogate keys and validity checks. This response fits a separate situation (A salesperson moves territories and historical sales must remain attributed to prior territory.), but does not resolve the stated case. The targeted response is: Preserve run versions, source snapshots and transformation lineage with approved restatement rules.
D:Provide certified datasets, semantic layer guidance and role-based BI access. This response fits a separate situation (Citizen analysts build dashboards using tables with unknown sensitivity and ambiguous joins.), but does not resolve the stated case. The targeted response is: Preserve run versions, source snapshots and transformation lineage with approved restatement rules.
Question 186
A data analyst must confirm the cause of this issue: A registration field is blank, but downstream calculations treat unknown age as zero. Which check provides the strongest diagnostic evidence?
- Inspect subtype constraints, shared keys and invalid category combinations.
- Review DDL, foreign keys, indexes and query plans against the logical model.
- Test enrollments, composite uniqueness and independent student/course updates.
- Inspect missing-value rules and test calculations with unknown and genuine zero.
Explanation
Correct Answer
D:Inspect missing-value rules and test calculations with unknown and genuine zero. This test checks the relevant business and technical failure rather than assuming its cause. Null can denote absent or unknown information and is not automatically a numeric zero. Expected control result: Unknown values remain distinguishable from legitimate zero and inappropriate imputations are flagged.
Incorrect Answers
A:Inspect subtype constraints, shared keys and invalid category combinations. This check would investigate another issue (A party table mixes person birthdate with organization registration number, leaving ambiguous rules.), not the immediate case. The direct diagnostic is: Inspect missing-value rules and test calculations with unknown and genuine zero.
B:Review DDL, foreign keys, indexes and query plans against the logical model. This check would investigate another issue (A logical model is approved but a PostgreSQL implementation needs columns and access paths.), not the immediate case. The direct diagnostic is: Inspect missing-value rules and test calculations with unknown and genuine zero.
C:Test enrollments, composite uniqueness and independent student/course updates. This check would investigate another issue (A student can enroll in many courses and a course can enroll many students.), not the immediate case. The direct diagnostic is: Inspect missing-value rules and test calculations with unknown and genuine zero.
Question 187
An accountable data owner wants both a response and a measurable check: An organization compares central warehouse, decentralized data products and a virtualization layer for analytics. Which action-and-evidence pair is BEST?
- Action: Assess patterns against consistency, latency, autonomy, cost, governance and security needs. Verification: The selected pattern has recorded tradeoffs and meets representative end-to-end workload tests.
- Action: Use a federated architecture with governed metadata and lawful cross-border access patterns. Verification: Queries respect localization constraints and approved shared views expose necessary aggregates.
- Action: Create agreed architecture principles and review exceptions through a decision process. Verification: Decision records show when a common pattern is followed and why exceptions are justified.
- Action: Agree enterprise-level business concepts and relationships before mapping physical application schemas. Verification: An approved logical model distinguishes party, agreement and financial account semantics.
Explanation
Correct Answer
A:Action: Assess patterns against consistency, latency, autonomy, cost, governance and security needs. Verification: The selected pattern has recorded tradeoffs and meets representative end-to-end workload tests. Both the intervention and measurable outcome are directly relevant. There is no universally best data architecture; fitness depends on measurable constraints.
Incorrect Answers
B:This plan addresses a different situation (A multinational has legal requirements to keep certain data local while enabling group analytics.), not the targeted problem. Here, action and evidence must match: Action: Assess patterns against consistency, latency, autonomy, cost, governance and security needs.
C:This plan addresses a different situation (Project teams repeatedly pick conflicting data integration styles for identical business needs.), not the targeted problem.
D:This plan addresses a different situation (Two departments use the term account for unrelated concepts and merge records incorrectly.), not the targeted problem.
Question 188
A data analyst must confirm the cause of this issue: A customer's billing address is validly formatted but refers to a previous residence. Which check provides the strongest diagnostic evidence?
- Trace missing IDs to upstream forms, interfaces and repeated source events.
- Review quality agreements, monitoring windows, breaches and corrective actions.
- Compare governed customer records with independently verified source evidence.
- Test form constraints, exception permissions and downstream defect rates.
Explanation
Correct Answer
C:Compare governed customer records with independently verified source evidence. This test checks the relevant business and technical failure rather than assuming its cause. Accuracy concerns correspondence to reality; valid formatting alone cannot prove accuracy. Expected control result: A sampled comparison confirms the stored address reflects current real-world facts.
Incorrect Answers
A:Trace missing IDs to upstream forms, interfaces and repeated source events. This check would investigate another issue (An analyst manually fixes the same missing supplier IDs every week.), not the immediate case. The direct diagnostic is: Compare governed customer records with independently verified source evidence.
B:Review quality agreements, monitoring windows, breaches and corrective actions. This check would investigate another issue (A shared customer dataset has no agreed defect response or availability expectations.), not the immediate case. The direct diagnostic is: Compare governed customer records with independently verified source evidence.
D:Test form constraints, exception permissions and downstream defect rates. This check would investigate another issue (Bad dates are entered during onboarding and corrected only after monthly reporting failures.), not the immediate case. The direct diagnostic is: Compare governed customer records with independently verified source evidence.
Question 189
In reviewing this data risk, select the LEAST defensible practice: An analyst manually fixes the same missing supplier IDs every week.
- Schedule more manual corrections without investigating why IDs go missing.
- Preserve source evidence and document why any exception was approved.
- Check relevant business definitions before joining independently sourced data.
- Document which dataset version was used in a material decision.
Explanation
Correct Answer
A:Schedule more manual corrections without investigating why IDs go missing. Sustainable data quality improvement addresses causes rather than repeated symptoms. A sound response instead: Identify the upstream source-process failure and implement a preventive correction.
Incorrect Answers
B:Preserve source evidence and document why any exception was approved. Traceability enables independent review of exceptions. The case-specific response is: Identify the upstream source-process failure and implement a preventive correction.
C:Check relevant business definitions before joining independently sourced data. Semantic agreements prevent incorrect cross-system correlations. The case-specific response is: Identify the upstream source-process failure and implement a preventive correction.
D:Document which dataset version was used in a material decision. Version records support reproducible decisions. The case-specific response is: Identify the upstream source-process failure and implement a preventive correction.
Question 190
After addressing this problem, what evidence would justify closure? A repository mixes invoices, procedures and complaints under inconsistent free-text tags.
- Search and lifecycle workflows retrieve the expected classes with measurable precision.
- Users find the currently approved policy and earlier revisions remain traceable.
- Search tests surface the current approved procedure rather than obsolete drafts.
- The record shows origin, integrity, retention rule and auditable access history.
Explanation
Correct Answer
A:It tests the action: Define a managed content taxonomy and classify documents using validated metadata.
Incorrect Answers
B:Users find the currently approved policy and earlier revisions remain traceable. This would verify a different control (Staff collaborate on a policy but several copied files are called final and no approved version is clear.), not the one in the question. The required evidence is: Search and lifecycle workflows retrieve the expected classes with measurable precision.
C:Search tests surface the current approved procedure rather than obsolete drafts. This would verify a different control (Employees cannot find approved procedures because files lack titles, tags and indexed text.), not the one in the question. The required evidence is: Search and lifecycle workflows retrieve the expected classes with measurable precision.
D:The record shows origin, integrity, retention rule and auditable access history. This would verify a different control (An archived approval email is needed as evidence in a dispute but its author and date are missing.), not the one in the question. The required evidence is: Search and lifecycle workflows retrieve the expected classes with measurable precision.