CDMP DQ

Data Quality (CDMP Definition)

The degree to which data meets business requirements and is fit for its intended use in operations, decision-making, and planning.

Data Quality Management (DQM)

A continuous lifecycle process that includes defining, measuring, analyzing, and improving data quality to align with organizational objectives.

Core Goals of Data Quality Management

1. Increase data trustworthiness and usability. 2. Reduce operational risk and costs caused by poor data. 3. Establish accountability and standards for data assets.

Data Quality Assessment (DQA)

The process of evaluating data against established quality dimensions and business rules using objective metrics.

Root Cause Analysis (RCA) in Data Quality

An analytical technique used to trace data errors back to their underlying technical, process, or human origins to prevent recurrence.

Accuracy (Data Quality Dimension)

The degree to which data correctly describes the real-world object, entity, or event it represents.

Completeness (Data Quality Dimension)

The proportion of required data that is present and recorded relative to the total expected set.

Consistency (Data Quality Dimension)

The absence of contradictions across different datasets, systems, or within a single dataset over time.

Timeliness (Data Quality Dimension)

The degree to which data is available at the expected time and frequency required for its intended business process.

Uniqueness (Data Quality Dimension)

The requirement that each real-world entity is represented only once within a dataset, ensuring the absence of duplicates.

Validity (Data Quality Dimension)

The degree to which data conforms to defined domain rules, syntax standards, formats, or allowed range values.

Scenario: A customer table contains age entries listed as $-5$, $250$, and $\text{"N/A"}$ in a numeric field bounded between $0$ and $120$. Which dimension is primarily violated?

Validity (the values fall outside the allowed domain and range rules).

Scenario: Two internal enterprise systems display conflicting shipping addresses for customer account ID $10492$ without any change audit record. Which dimension is primarily violated?

Consistency (data differs across systems without a clear single point of truth).

Scenario: A regulatory report requires mandatory fields like Tax_ID and Legal_Name, but $35\text{\%}$ of business records have these fields left null. Which dimension is primarily violated?

Completeness (expected mandatory data elements are missing).

Scenario: A real-time automated trading platform receives stock market feeds with a $10$-minute delay, causing suboptimal automated trades. Which dimension is primarily violated?

Timeliness (data is not available at the required frequency or velocity for real-time operations).

Scenario: A marketing software sends four identical emails to one individual because they exist as four separate records with minor name variations. Which dimension is primarily violated?

Uniqueness (failure to consolidate multiple representations of the same real-world entity).

Data Quality (CDMP Definition)

The degree to which data meets business requirements and is fit for its intended use in operations, decision-making, and planning.

Data Quality Management (DQM)

A continuous lifecycle process that includes defining, measuring, analyzing, and improving data quality to align with organizational objectives.

Core Goals of Data Quality Management

1. Increase data trustworthiness and usability. 2. Reduce operational risk and costs caused by poor data. 3. Establish accountability and standards for data assets.

Data Quality Assessment (DQA)

The process of evaluating data against established quality dimensions and business rules using objective metrics.

Root Cause Analysis (RCA) in Data Quality

An analytical technique used to trace data errors back to their underlying technical, process, or human origins to prevent recurrence.

Accuracy (Data Quality Dimension)

The degree to which data correctly describes the real-world object, entity, or event it represents.

Completeness (Data Quality Dimension)

The proportion of required data that is present and recorded relative to the total expected set.

Consistency (Data Quality Dimension)

The absence of contradictions across different datasets, systems, or within a single dataset over time.

Timeliness (Data Quality Dimension)

The degree to which data is available at the expected time and frequency required for its intended business process.

Uniqueness (Data Quality Dimension)

The requirement that each real-world entity is represented only once within a dataset, ensuring the absence of duplicates.

Validity (Data Quality Dimension)

The degree to which data conforms to defined domain rules, syntax standards, formats, or allowed range values.

Scenario: A customer table contains age entries listed as $-5$, $250$, and $\text{"N/A"}$ in a numeric field bounded between $0$ and $120$. Which dimension is primarily violated?

Validity (the values fall outside the allowed domain and range rules).

Scenario: Two internal enterprise systems display conflicting shipping addresses for customer account ID $10492$ without any change audit record. Which dimension is primarily violated?

Consistency (data differs across systems without a clear single point of truth).

Scenario: A regulatory report requires mandatory fields like Tax_ID and Legal_Name, but $35\text{\%}$ of business records have these fields left null. Which dimension is primarily violated?

Completeness (expected mandatory data elements are missing).

Scenario: A real-time automated trading platform receives stock market feeds with a $10$-minute delay, causing suboptimal automated trades. Which dimension is primarily violated?

Timeliness (data is not available at the required frequency or velocity for real-time operations).

Scenario: A marketing software sends four identical emails to one individual because they exist as four separate records with minor name variations. Which dimension is primarily violated?

Uniqueness (failure to consolidate multiple representations of the same real-world entity).

Exam Question: What is the primary difference between Data Profiling and Data Quality Assessment (DQA)?

Data Profiling discovers data structure, statistics, and anomalies without pre-defined rules, whereas DQA evaluates data against explicit, pre-defined business rules.

Exam Question: Which Data Governance role is explicitly responsible for defining data quality business rules and thresholds?

Data Steward (or Business Data Steward).

Exam Question: What process identifies and merges disparate customer records representing the same real-world individual?

Entity Resolution (or Record Matching / Deduplication).

Exam Question: If a completeness metric calculation formula is $\frac{\text{Populated Fields}}{\text{Total Required Fields}} \times 100\%$, what is the result when $80$ out of $100$ required fields are populated?

$80\%$ (calculated as $\frac{80}{100} \times 100\% = 80\%$).

Exam Question: What term describes the delay between when a real-world event occurs and when its corresponding data becomes available in an operational system?

Data Latency.

Exam Question: What formal agreement defines expected data quality standards, thresholds, and remediation response times between producers and consumers?

Data Quality Service Level Agreement (SLA).