Free DAMA CDMP Quality Practice Questions: Profiling Discovery and Assessment

Topic snapshot

Sample questions

These are original IT Mastery practice questions aligned to this topic area. They are not official exam questions, copied live-exam content, or exam dumps. Use them for self-assessment, scope review, and deciding what to drill next.

Question 1

Topic: Profiling Discovery and Assessment

A bank has newly designated customer_tax_id as a critical data element. There is no prior quality measurement. An initial profiling run over 12 months found 8,430 invalid values, and product owners want a starting point for trend reporting before setting ongoing alert levels. Which action best fits the need?

Options:

• A. Document the profiled validity rate, scope, date, and method as the baseline.

• B. Use 8,430 invalid values as the quality baseline.

• C. Set an alert when invalid values exceed 1% in any daily load.

• D. Close the issue after the invalid values are corrected once.

Best answer: A

Explanation: A data quality baseline is the measured starting point for a data element, rule, dimension, population, and time period. It should include enough context to make future comparisons meaningful, such as the profiling method, population covered, measurement date, and observed rate. The count of 8,430 invalid records is useful evidence from the assessment, but by itself it is a one-time exception count and may not support fair trend comparison if volumes change. An ongoing control threshold is different: it defines when monitoring should alert, escalate, or fail a process after acceptable tolerance has been agreed. The bank needs the initial measured state before it can judge trends or set operational thresholds.

• Daily alert threshold is an ongoing control decision, not the initial measured state requested by the product owners.
• Exception count only is incomplete because it lacks rate, scope, and method for future comparison.
• One-time correction treats the defect as a cleanup task and does not establish measurement for trend reporting.

Question 2

Topic: Profiling Discovery and Assessment

A data quality analyst profiles 50,000 records in an order history dataset. The business rule says a shipped order must have order_date on or before ship_date.

Which quality issue is most supported by the profiling summary?

Options:

• A. Reference data validity defect

• B. Cross-field consistency rule violation

• C. Completeness defect in shipment dates

• D. Duplicate order identifier defect

Best answer: B

Explanation: The profile most strongly supports a cross-field consistency, or integrity, issue. Each individual date field is populated and parses successfully, so the basic completeness and format-validity checks pass. The defect appears only when the two fields are evaluated together against the business rule: a shipped order should not have a shipment date earlier than its order date. Profiling is useful here because it separates single-column findings from relationship-based rule failures. The decisive evidence is the count of shipped records where ship_date precedes order_date, not the presence of missing values, invalid codes, or duplicate identifiers.

• Completeness does not fit because the profile shows no missing shipment dates for shipped orders.
• Reference data validity is not supported because no code set, lookup table, or allowed-value comparison is shown.
• Duplicate identifiers are ruled out by the profile result showing zero duplicate order_id values.

Question 3

Topic: Profiling Discovery and Assessment

A bank’s monthly customer profitability report is overstating revenue for a subset of small-business customers. Profiling shows valid customer IDs and complete revenue amounts in the billing source. The defect appears only after data is prepared for the data warehouse.

Exhibit: Source-to-consumption evidence

Where is the quality defect most likely introduced or amplified?

Options:

• A. At the ETL join to CRM master data

• B. In the BI aggregation formula

• C. In the billing source invoice capture

• D. In the customer ID format validation rule

Best answer: A

Explanation: The evidence points to a lifecycle defect introduced during integration, not at initial capture. Billing has complete amounts and valid customer IDs, so the source record quality is not the main failure. The CRM master data contains duplicate identities for acquired customers, but the overstatement occurs when the ETL process joins invoice rows to both identifiers. That join multiplies otherwise valid invoice records, and the BI layer then aggregates the duplicated rows by enterprise customer.

The key distinction is between the underlying master data weakness and the point where it becomes a measured business defect. Here, the defect is amplified in staging because the integration rule does not handle duplicate or merged customer identifiers.

• Billing capture is unlikely because profiling found valid customer IDs and complete revenue amounts at the source.
• BI aggregation consumes duplicated prepared data; it does not explain why invoice rows were duplicated before reporting.
• Format validation would check syntactic validity, but the issue is identity duplication and join cardinality.

Question 4

Topic: Profiling Discovery and Assessment

A billing analytics team finds missing tax identifiers in customer records used for regulatory invoices. The CRM source team says the identifiers are complete in their application, while the warehouse team says the defect appears before reporting. Current quality rules are informal, and only the warehouse has been profiled. What is the best discovery response?

Options:

• A. Cleanse the warehouse records and close the issue log

• B. Require the CRM team to fix the missing identifiers immediately

• C. Profile each lifecycle handoff and review lineage with accountable stewards

• D. Change the billing report to exclude records without identifiers

Best answer: C

Explanation: When downstream teams see defects and source teams dispute responsibility, discovery should first establish where the defect is introduced or transformed. DAMA-aligned data quality practice uses profiling, lineage, and source process discovery to connect observed defects to lifecycle events, controls, and accountable stewardship roles. In this case, only the warehouse has been profiled, so assigning blame to CRM or cleaning only the warehouse would be premature. A joint review of handoffs, mappings, extracts, transformations, and rule definitions creates shared evidence and supports sustainable remediation. The key is to move from disputed symptoms to traceable cause and ownership.

• Immediate CRM fix assumes source responsibility before evidence shows where the missing values arise.
• Warehouse cleansing treats the downstream symptom but does not prevent recurrence or resolve ownership.
• Report exclusion hides business impact and may worsen regulatory invoice completeness.

Question 5

Topic: Profiling Discovery and Assessment

A data quality analyst profiles customer onboarding data before publishing a scorecard. The business steward says a finding becomes a quality defect only when it violates an approved business rule or reduces fitness for a stated business use. Which profiling finding should be logged as a business-relevant quality defect rather than only a statistically interesting anomaly?

Options:

• A. Middle initials are blank for 72% of customers

• B. A spike in signups every Friday afternoon

• C. VIP customers missing consent status used for campaigns

• D. Postal codes starting with 9 are unusually frequent

Best answer: C

Explanation: A statistically interesting anomaly is a pattern that stands out in profiling, such as an unusual frequency, spike, or distribution. It becomes a data quality defect only when business context shows that it violates a defined rule, threshold, or fitness-for-purpose requirement. In this scenario, consent status is needed for campaign use, and missing values for VIP customers directly affect business action and compliance-sensitive decisioning. The other findings may deserve investigation, but the stem gives no approved rule or business use that they break. The key distinction is not statistical surprise; it is business impact against an agreed quality expectation.

• Friday spike may be seasonality or process timing; no rule or business harm is stated.
• Blank middle initials may be acceptable if the attribute is optional or unused.
• Frequent 9-prefix postal codes is a distribution pattern, but no invalidity or business impact is shown.

Question 6

Topic: Profiling Discovery and Assessment

A data quality analyst profiles a customer onboarding file before loading it to the customer master. The profile shows unexpected customer_status values, missing records from one region, duplicate tax identifiers, and three different date formats. The source team says these patterns may reflect local onboarding practices, but no approved quality rules exist yet. What is the best next step?

Options:

• A. Reject the full file until all formats match

• B. Build a recurring scorecard from the current profile

• C. Confirm business rules with the steward and prioritize issues

• D. Cleanse all invalid-looking values in the staging area

Best answer: C

Explanation: Profiling discovers patterns, anomalies, and potential defects; it does not by itself determine business fitness for purpose. Here, the analyst has found several possible quality issues, but local practices may explain some values or formats, and no approved quality rules exist. The best next step is to involve the accountable steward or business owner to confirm definitions, valid values, required records, uniqueness expectations, and impact. That turns profiling observations into agreed quality rules, thresholds, and prioritized issues. Cleansing, rejection, or scorecarding should follow from confirmed rules and business impact, not from assumptions based only on technical appearance.

• Immediate cleansing fails because values that look invalid may be legitimate local codes until business rules are confirmed.
• Premature scorecarding fails because recurring measures need approved rules, thresholds, and ownership.
• Full-file rejection fails because it applies a control before assessing which findings are true defects and how severe they are.

Question 7

Topic: Profiling Discovery and Assessment

A data quality assessment finds exceptions in the monthly invoice dataset. Which action best fits the evidence?

Options:

• A. Refine the rule to exclude discounted orders

• B. Profile more months before taking action

• C. Escalate the rule for stakeholder review

• D. Remediate the integration mapping and reload affected invoices

Best answer: D

Explanation: Assessment evidence supports immediate remediation when the defect is confirmed, the business rule is still valid, the threshold is breached, and the root cause is sufficiently isolated. Here, the failures began after a known mapping change, the source orders contain the correct tax values, and finance confirms that the invoice rule and threshold remain appropriate. That points to a process defect in the integration mapping, not a questionable rule or an unresolved business definition. The sustainable response is to correct the mapping and reload or repair the affected invoices, with normal issue tracking and verification after the fix. Further profiling or stakeholder review would delay action without adding necessary decision evidence.

• More profiling is unnecessary because the failure pattern, root cause, and business impact are already clear enough to act.
• Rule refinement fails because finance confirmed the rule applies; discounted orders are affected by a mapping defect, not a valid exception.
• Stakeholder review is not the next step because the business owner has already validated the rule and threshold.

Question 8

Topic: Profiling Discovery and Assessment

During a baseline data quality assessment at a bank, the governance group can assign enhanced rules, monitoring, ownership, and remediation capacity to only one data element this quarter.

Which decision best fits the assessment evidence?

Options:

• A. Prioritize loan maturity date for enhanced quality management

• B. Prioritize customer mobile phone because it has the highest blank rate

• C. Prioritize customer nickname because completeness is lowest

• D. Prioritize branch code because obsolete values still appear

Best answer: A

Explanation: A critical data element should receive stronger quality attention when poor quality could materially affect business operations, regulatory reporting, customer outcomes, or other high-value decisions. The loan maturity date supports both payment scheduling and a regulatory liquidity report, so even a smaller defect rate can be more important than a larger defect rate on a low-impact field. The profiling evidence also shows a clear validity problem and an ownership gap, which makes stronger rules, monitoring, stewardship assignment, and remediation planning appropriate. High defect volume alone does not make a field critical; fitness for purpose and business impact drive prioritization.

• Highest blank rate is tempting, but mobile phone is optional for most products and mainly affects marketing use.
• Obsolete branch codes are already handled by a nightly remapping control, reducing the need for new priority treatment.
• Low nickname completeness has limited business impact because the field is used only for online display.

Question 9

Topic: Profiling Discovery and Assessment

A data steward receives complaints that the current rule, “customer email must be populated,” is producing too many exceptions for records loaded from a legacy call-center source. Business users suggest changing the rule to apply only to online customers. Before changing the rule, what evidence is most appropriate to collect through profiling?

Options:

• A. A revised exception threshold approved by the data quality team

• B. A one-time cleanup list for all missing email values

• C. Completeness and usage patterns by source and customer channel

• D. A new scorecard showing fewer email exceptions

Best answer: C

Explanation: Profiling supports rule definition by discovering actual data patterns before a quality rule is created or changed. In this case, the proposed rule change depends on whether missing email values are concentrated in a source, associated with a customer channel, and meaningful for business use. The strongest evidence is a profile that segments completeness results by source system and customer channel, ideally paired with usage context from stakeholders. That evidence helps distinguish a valid business exception from a recurring defect or source-process issue. Changing thresholds, cleansing records, or adjusting scorecards before understanding the pattern can hide quality problems rather than improve fitness for purpose.

• Threshold approval skips the discovery needed to know whether the proposed exception is justified.
• One-time cleanup treats symptoms but does not determine whether the rule itself fits the data and business need.
• Scorecard reduction improves reported results cosmetically if the underlying pattern and rule rationale are not understood.

Question 10

Topic: Profiling Discovery and Assessment

A data steward is preparing a customer data quality plan for a new integration feed. Initial sampling shows inconsistent phone formats, unexpected nulls in date_of_birth, and several duplicate-looking customer records. No approved business rules or thresholds exist yet, but the feed will later support fraud screening. What is the best next data quality activity?

Options:

• A. Profile the feed to discover patterns and candidate rules

• B. Reject records that violate proposed fraud-screening rules

• C. Run automated remediation to merge duplicate customers

• D. Publish a compliance scorecard against formal thresholds

Best answer: A

Explanation: Data profiling is used for discovery and assessment when rules are not yet mature. It examines distributions, formats, null rates, duplication patterns, outliers, and relationships so stewards and business owners can understand the data and define quality expectations. Validation is different: it checks data against approved rules or controls, such as required fields, valid codes, or threshold limits. Remediation is execution work that corrects confirmed defects, such as standardizing values or resolving duplicates, usually after rules, ownership, and treatment decisions are clear. In this case, the feed has visible anomalies and important downstream use, but no approved rules or thresholds, so discovery should come before control enforcement or cleanup.

• Premature rejection fails because fraud-screening rules have not been approved for validation enforcement.
• Automated merging is risky because duplicate-looking records require confirmed matching rules and stewardship decisions.
• Scorecard publishing fails because formal thresholds do not yet exist for meaningful control reporting.

Continue in the web app

Use IT Mastery for interactive DAMA CDMP Data Quality Specialist practice with mixed sets, timed mocks, topic drills, explanations, and progress tracking.

Try DAMA CDMP Data Quality Specialist on Web

Related focused pages

• Free DAMA CDMP Data Quality Specialist Full-Length Practice Exam
• Data Quality Foundations and Business Fitness
• Quality Strategy Business Case and Prioritization
• Quality Rules Standards and Requirements
• Root Cause Analysis and Remediation
• Monitoring Scorecards and Measurement
• Quality Governance Roles and Stewardship
• Metadata Lineage and Quality Evidence
• Master Reference and Warehouse Quality
• Quality in Data Lifecycle and Operations
• Quality Maturity and Continuous Improvement
• Specialist Cross-Discipline Judgment