2 packs: 15% off with code BUY2. 3 or more: 20% off with code BUY3.

Free DAMA CDMP Data Management Fundamentals Practice Questions

36 free DAMA CDMP Data Management Fundamentals practice questions from the CertStash pack of 295, each with the correct answer and a full explanation. The rule is the same for every exam: ten free on the product page, 36 free here, all 295 in the pack, and each step includes the one before it.

Get all 295 questions (US$39) · Download the first 10 as a PDF

Question 1

The best way to validate that a database backup is working, is to:

  1. Periodically recover from the backup file
  2. Check the backup logs every day
  3. Appoint a special DBA in charge of backups
  4. Check the size of the backup file
  5. Check for the automatic email notification of backup success
Show answer and explanation

Correct answer: A. Periodically recover from the backup file

A backup is only proven useful when data can actually be restored from it. DMBOK's Data Storage and Operations guidance calls for regularly testing recovery, ideally through practice restores onto non-production copies or failover systems, so that problems with backup files, procedures or media are found before a real outage. Logs, file sizes and success notifications only show that a backup job ran, not that the resulting file is complete and restorable. Periodically recovering from the backup is therefore the best way to validate it.

Why the other options are wrong

  • B. Backup logs confirm that a job ran, not that its output can be restored.
  • C. Assigning a dedicated DBA adds accountability but does not test whether the backup actually works.
  • D. A plausible file size says nothing about whether the contents are complete or restorable.
  • E. A success notification only reports that the job finished, not that recovery from it will succeed.

Question 2

When reviewing data access plans, sequential searching is slowing the database.

One way to fix this is:

  1. Reducing the number of database users
  2. Creating new indexes
  3. Converting it to an in-memory database
  4. Moving the database to the cloud
  5. Adding more memory
Show answer and explanation

Correct answer: B. Creating new indexes

Sequential searching means the database is scanning tables row by row because no suitable access path exists for the query. When a review of data access plans reveals full scans on large tables, the standard tuning response in database performance management is to create indexes on the columns used for selection and joins. An index lets the optimizer go directly to the matching rows, which removes the scan and speeds up the query without changing hardware, platform or the user base.

Why the other options are wrong

  • A. Having fewer users reduces load but leaves the inefficient scans in every query.
  • C. Converting to an in-memory database is a major platform change and does not address the missing access path.
  • D. Cloud hosting changes the infrastructure, not the query's access plan.
  • E. Extra memory may cache more data, but the queries still scan sequentially.

Question 3

What area do you not consider when developing a 'data governance operating model'?

  1. The availability of industry data models
  2. The business model (decentralized versus centralized)
  3. The value of data to the organisation
  4. Cultural factors - such as acceptance of discipline and adaptability to change
  5. Impact of regulation
Show answer and explanation

Correct answer: A. The availability of industry data models

When designing a data governance operating model, DMBOK identifies the factors that shape it: the value of data to the organization, the business model (decentralized or centralized, local or international), cultural factors such as acceptance of discipline and adaptability to change, and the impact of regulation. These determine how centralized the model should be, how much formal structure people will accept and what compliance demands it must meet. The availability of industry data models is a data modeling and architecture resource, not a factor in designing the governance operating model.

Why the other options are wrong

  • B. The centralized or decentralized business model is a core input to the operating model design.
  • C. The value of data to the organization is a key factor in how much governance is warranted.
  • D. Cultural factors determine how readily the organization accepts governance discipline and change, so they are considered.
  • E. Regulation strongly shapes the governance model, especially in heavily regulated industries.

Question 4

The purpose of data governance is to ensure that:

  1. Data can be reported on by the business line
  2. Data is backed up every night
  3. Data will be understood by all stakeholders
  4. Data is available for other systems to use
  5. Data is managed properly, according to policies and best practices
Show answer and explanation

Correct answer: E. Data is managed properly, according to policies and best practices

DMBOK states that the purpose of data governance is to ensure that data is managed properly, according to policies and best practices. Governance is the exercise of authority, control and shared decision making over the management of data assets: it sets policies, standards and decision rights, and oversees compliance with them. Outcomes such as reporting, availability, shared understanding and backup are results of good data management in specific knowledge areas, while governance exists to make sure all data management activity follows agreed policies and practices.

Why the other options are wrong

  • A. Business line reporting is a use of data, not the purpose of governance.
  • B. Nightly backups are an operational task within Data Storage and Operations.
  • C. Shared understanding is a benefit supported by metadata and definitions, not the overall purpose of governance.
  • D. Making data available to other systems is an integration concern, not the purpose of governance.

Question 5

A common driver for data governance is:

  1. Irreconcilable figures in reports
  2. Regulatory compliance
  3. The appointment of a CDO
  4. Internal audits
  5. Outsourcing
Show answer and explanation

Correct answer: B. Regulatory compliance

Regulatory compliance is the most common driver for data governance, particularly in heavily regulated industries such as financial services and healthcare. Laws and regulations require organizations to show how data is controlled, protected and reported, and responding to these obligations often provides the initial mandate and funding for a governance program. Over time other drivers such as analytics, risk reduction and process improvement grow in importance, but compliance remains the classic starting point.

Why the other options are wrong

  • A. Inconsistent report figures are a symptom that governance addresses, not the common driver.
  • C. Appointing a CDO is usually a response to the need for governance, not its driver.
  • D. Internal audits may surface data problems but are not the typical driver.
  • E. Outsourcing is not a recognized common driver for data governance.

Question 6

Adoption of a Data Governance program is most likely to succeed:

  1. When dictated by senior executives
  2. When the entire enterprise is partaking at once
  3. In 1 or 2 months with a large consulting team
  4. When the CDO is a charismatic leader
  5. With an incremental rollout strategy
Show answer and explanation

Correct answer: E. With an incremental rollout strategy

Data governance requires changes in behavior and culture, so DMBOK recommends implementing it incrementally rather than across the whole enterprise at once. A phased rollout starts with a limited scope, such as one business area or a high-value data problem, demonstrates value, learns from experience and then expands. This builds sponsorship and acceptance gradually and lets the program adjust its policies and processes as it grows. Organizational change of this kind takes sustained effort over time, not a mandate or a short consulting engagement.

Why the other options are wrong

  • A. A mandate from executives without gradual adoption tends to meet resistance and does not create lasting change.
  • B. Engaging the whole enterprise at once is too much change at one time and raises the risk of failure.
  • C. Governance is a long-term cultural change and cannot be established in one or two months by consultants.
  • D. Personal charisma does not replace a structured, sustained implementation approach.

Question 7

In 2009, ARMA International published GARP for managing records and information.

GARP stands for:

  1. Generally Accepted Recordkeeping Principles
  2. Generally Available Recordkeeping Practices
  3. Gregarious Archive of Recordkeeping Processes
  4. Global Accredited Recordkeeping Principles
  5. G20 Approved Recordkeeping Principles
Show answer and explanation

Correct answer: A. Generally Accepted Recordkeeping Principles

GARP stands for Generally Accepted Recordkeeping Principles, issued by ARMA International in 2009. The eight principles (accountability, integrity, protection, compliance, availability, retention, disposition and transparency) describe how business records should be maintained and provide a framework for recordkeeping and information governance, with an associated maturity model for assessing an organization's practice. DMBOK uses GARP within Document and Content Management to assess records management maturity.

Why the other options are wrong

  • B. The principles are generally accepted, and they are principles rather than practices.
  • C. This is not what the GARP acronym stands for.
  • D. GARP is not described as global or accredited.
  • E. GARP has no connection to the G20.

Question 8

'Planning, implementation and control activities for lifecycle management of data and information, found in any form or medium', pertains to which knowledge area?

  1. Data Warehousing and Business Intelligence
  2. Data Integration and Interoperability
  3. Metadata Management
  4. Document and Content Management
  5. Data Storage and Operations
Show answer and explanation

Correct answer: D. Document and Content Management

DMBOK defines Document and Content Management as the planning, implementation and control activities for lifecycle management of data and information found in any form or medium, with particular emphasis on documents and records needed for legal and regulatory compliance. It covers the capture, storage, access, retention and disposal of unstructured content such as documents, records, email and multimedia. The phrase "in any form or medium" is the distinguishing feature, because the other knowledge areas focus mainly on structured data.

Why the other options are wrong

  • A. Data Warehousing and Business Intelligence covers decision support data and reporting.
  • B. Data Integration and Interoperability covers the movement and consolidation of data between systems.
  • C. Metadata Management covers the data about data that enables management of other data.
  • E. Data Storage and Operations covers the design, implementation and support of stored data, mainly in databases.

Question 9

Which of the following is a reason why organisations do not dispose of non-value-adding information?

  1. The organisation's data quality benchmark diminishes
  2. Data modelling the content is hard to reproduce
  3. The information is never out of date
  4. Storage is cheap and easily expanded
  5. The metadata repository cannot be updated
Show answer and explanation

Correct answer: D. Storage is cheap and easily expanded

DMBOK lists several reasons organizations fail to dispose of non-value-adding information: inadequate policies, disagreement about which information has value, inability to foresee future needs, lack of buy-in for records management, inability to decide what to delete, and the perceived cost of making disposal decisions. Another is that electronic storage is cheap and easy to expand, so buying more space seems simpler than archiving and removal. Retaining such information still carries cost and legal risk, because it remains discoverable in litigation.

Why the other options are wrong

  • A. Disposing of non-value-adding information does not lower any data quality benchmark.
  • B. Difficulty reproducing data models is not a reason for keeping unneeded information.
  • C. Information does go out of date, which is why retention schedules and disposal exist.
  • E. Disposal is not blocked by any inability to update a metadata repository.

Question 10

An enterprise content management system (ECM) stores the following:

  1. Kanban boards
  2. Backups of the enterprise's data warehouse
  3. Code and test scripts for enterprise development environments
  4. Documents and images, but not multi-media
  5. Information to be displayed on websites and contained in documents
Show answer and explanation

Correct answer: E. Information to be displayed on websites and contained in documents

An enterprise content management system captures, stores, manages and delivers the organization's content: the information displayed on websites and contained in documents. DMBOK describes content as the data and information inside a file, document or website, and content management as organizing, categorizing and structuring those resources so they can be stored, delivered and reused in many ways. ECM therefore supports documents, web content and records across the enterprise, typically including multimedia such as audio and video.

Why the other options are wrong

  • A. Kanban boards live in project and work management tools.
  • B. Data warehouse backups are handled by database backup and recovery processes.
  • C. Code and test scripts belong in source control and development tooling.
  • D. ECM systems manage multimedia content as well as documents.

Question 11

Is the risk of a data breach higher in documents on a network drive or in an enterprise resource planning system (ERP)?

  1. Data needs to be emailed to be breached effectively
  2. The probability is exactly the same
  3. Data cannot be breached from either platform due to the superior security in modern computer networks
  4. On a network drive
  5. In an enterprise resource planning system (ERP)
Show answer and explanation

Correct answer: D. On a network drive

An enterprise system such as an ERP holds data in a managed database with defined user roles, application-level authorization and audit trails. Documents on a network drive are files that can be copied, forwarded and saved elsewhere with far less control, often with broad folder permissions and little monitoring of who opens them. DMBOK notes that the more places confidential data is stored, the higher the risk of a breach, and loosely controlled file copies multiply those places. The risk is therefore higher for documents on a network drive.

Why the other options are wrong

  • A. Data can be breached in many ways besides email.
  • B. The two platforms have very different levels of access control and auditing, so the risk is not the same.
  • C. No platform is immune to breaches, however modern its network security.
  • E. An ERP enforces role-based access and auditing that loosely controlled files lack.

Question 12

A goal of reference and master data management is to ensure shared data is:

  1. Secure, auditable, complete and relevant
  2. Secure, auditable, publicly available and free
  3. Continuous, consistent, current and private
  4. Complete, consistent, current and authoritative
  5. Complete, consistent, content and relevant
Show answer and explanation

Correct answer: D. Complete, consistent, current and authoritative

A goal of Reference and Master Data Management is to ensure the organization has complete, consistent, current and authoritative master and reference data across its processes. Other goals are enabling these data to be shared across business units and applications, and lowering cost and complexity through standards, common data models and integration patterns. Shared data with these four qualities can be trusted wherever it is used.

Why the other options are wrong

  • A. Security and auditability are Data Security concerns, and the stated goal calls for consistent, current and authoritative data.
  • B. Master data is not intended to be publicly available or free.
  • C. Continuous and private are not among the stated qualities.
  • E. Content is not a data quality, and the stated goal calls for authoritative data.

Question 13

A 'Content Distribution Network' supporting a multi-national website is likely to use:

  1. A database backup and restore solution
  2. An extract transform and load solution
  3. A records disposal solution
  4. A replication solution
  5. An archiving solution
Show answer and explanation

Correct answer: D. A replication solution

A content distribution network places copies of web content on servers in many geographic locations so that users are served from a nearby node, giving faster response and higher availability. Keeping those copies current means content must be copied from the origin to every node and kept synchronized as it changes. That is a replication solution: the same data maintained in multiple locations, which DMBOK describes under Data Storage and Operations as a way to improve availability and performance.

Why the other options are wrong

  • A. Backup and restore protects against loss but does not distribute content to users.
  • B. ETL transforms and loads data for integration or warehousing, not for serving web content.
  • C. Records disposal removes information at the end of its life and has nothing to do with distribution.
  • E. Archiving moves inactive content to long-term storage rather than delivering active content.

Question 14

Data flows can be represented by:

  1. Two dimensional matrices showing the relationships between data users and business processes
  2. Two dimensional matrices showing the relationships between data entities and business processes
  3. Two dimensional matrices showing the relationships between data facts and data dimensions
  4. Two dimensional matrices showing the relationships between data entities and data quality
  5. Two dimensional matrices showing the relationships between data users and data consumers
Show answer and explanation

Correct answer: B. Two dimensional matrices showing the relationships between data entities and business processes

DMBOK describes data flows as a form of data lineage documentation showing how data moves through business processes and systems. Besides diagrams, data flows can be represented as two-dimensional matrices that relate data entities to business processes, showing which processes create, read, update or delete which data. These matrices, similar to CRUD matrices, help architects see dependencies and gaps between the data and the processes that use it.

Why the other options are wrong

  • A. Relating data users to business processes describes roles, not the flow of data.
  • C. Facts and dimensions describe a dimensional warehouse model, not data flows.
  • D. Relating entities to data quality is a quality assessment, not a flow representation.
  • E. Data users and data consumers are overlapping groups of people, so relating them shows no data movement.

Question 15

Database performance depends upon two independent facets. These are:

  1. Availability and speed
  2. Choice of DBMS and programming language
  3. Hardware and network
  4. Distance to data center and network bandwidth
  5. Number of users and number of tables
Show answer and explanation

Correct answer: A. Availability and speed

DMBOK describes database performance as depending on two facets: availability and speed. Availability is the percentage of time the database can be used for productive work, and speed is how quickly and efficiently it returns data. A database that is fast but often unavailable, or available but slow, does not perform well, so DBAs manage both through tuning, monitoring, capacity planning and service level agreements.

Why the other options are wrong

  • B. DBMS choice and programming language are design decisions, not the two facets of performance.
  • C. Hardware and network influence performance but are not the two facets DMBOK defines.
  • D. Distance and bandwidth affect response times but are not the defined facets.
  • E. User and table counts are workload factors, not facets of performance.

Question 16

The goals of data storage and operations are:

  1. Managing accessibility, ensuring authorized access and managing performance
  2. Managing availability, ensuring consistency and managing glossaries
  3. Managing user experience, ensuring consistency and managing glossaries
  4. Managing user experience, ensuring data access and managing data lifecycles
  5. Managing availability, ensuring integrity and managing performance
Show answer and explanation

Correct answer: E. Managing availability, ensuring integrity and managing performance

DMBOK defines three goals for Data Storage and Operations: manage the availability of data throughout the data lifecycle, ensure the integrity of data assets, and manage the performance of data transactions. Together they focus the knowledge area on keeping data accessible when needed, protected from corruption or loss, and delivered efficiently, through activities such as database support, backup and recovery, and performance tuning.

Why the other options are wrong

  • A. Authorized access is a Data Security goal, and the stated goals refer to availability and integrity.
  • B. Managing glossaries belongs to Metadata Management and Data Governance.
  • C. User experience and glossaries are not goals of this knowledge area.
  • D. User experience is not a stated goal, and the list omits integrity and performance.

Question 17

In a data model, domains can be used for:

  1. Reducing the amount of coding required in the database
  2. Public viewing of the data model
  3. Consolidation of data into subject areas
  4. Consistency and standardization of data values
  5. Separating concepts into business ideas
Show answer and explanation

Correct answer: D. Consistency and standardization of data values

In data modeling, a domain is the complete set of possible values that an attribute can be assigned, defined by a data type, format, range, list of valid values or rule. Assigning attributes to shared domains, such as a standard date format or a currency code list, means every attribute using the domain is constrained to the same valid values. This promotes consistency and standardization of data values across the model and supports data quality checks.

Why the other options are wrong

  • A. Domains define valid values; they are not a means of reducing database code.
  • B. Domains govern attribute values and have nothing to do with who can view the model.
  • C. Grouping data into subject areas is done with subject area models, not domains.
  • E. Separating concepts into business ideas is the role of conceptual modeling.

Question 18

In order to gain endorsement to extend the use of reference and master data across the enterprise, a key supporting DMBOK knowledge area is:

  1. Data Architecture
  2. Data Governance
  3. Data Storage and Operations
  4. Data Security
  5. Document and Content Management
Show answer and explanation

Correct answer: B. Data Governance

Extending reference and master data across the enterprise requires agreement between business areas on definitions, ownership, standards and which sources are authoritative. Data Governance provides the decision rights, policies, stewardship and executive sponsorship needed to reach and enforce those agreements. DMBOK stresses that master data management is as much an organizational challenge as a technical one, so governance is the key supporting knowledge area for gaining endorsement and adoption across the enterprise.

Why the other options are wrong

  • A. Data Architecture designs where master data lives, but endorsement comes through governance decision making.
  • C. Data Storage and Operations supports the databases technically, not organizational endorsement.
  • D. Data Security protects master data but does not secure enterprise agreement on its use.
  • E. Document and Content Management deals with unstructured content, not shared master data.

Question 19

A data architect is best deployed:

  1. To build the database solution
  2. To manage the delivery of all the data aspects of a program
  3. After the project completes to identify weaknesses and lessons learned
  4. During the early stages of a project to define and shape a strategic solution
  5. By project managers to sign off all data deliverables
Show answer and explanation

Correct answer: D. During the early stages of a project to define and shape a strategic solution

A data architect provides the most value early in a project, when requirements are understood and the solution is being shaped. At that stage the architect can align the design with the enterprise data architecture, identify reuse of existing data and models, set standards and steer the project toward a strategic rather than tactical solution. Involvement after key design decisions have been made limits the architect to finding problems that are expensive to fix, so early engagement is where the role is best deployed.

Why the other options are wrong

  • A. Building the database is the work of developers and database administrators, not the architect's primary contribution.
  • B. Managing delivery of a program's data aspects is a project or delivery management role.
  • C. Reviewing weaknesses after completion is too late to shape the solution.
  • E. Signing off deliverables is a governance or quality step, not the best use of architectural skills.

Question 20

A project scope includes the collection, exchange and reporting of data from multiple systems. Conceptual, logical and physical data models are maintained.

How many models of each type can be expected?

  1. More logical data models than physical data models, and more logical data models than conceptual data models
  2. The same number of each of the model types
  3. More conceptual data models than logical data models, and more logical data models than physical data models
  4. Only 1 conceptual data model, 1 logical data model and 1 physical data model
  5. More physical data models than logical data models, and more logical data models than conceptual data models
Show answer and explanation

Correct answer: E. More physical data models than logical data models, and more logical data models than conceptual data models

Conceptual, logical and physical models describe data at increasing levels of detail. A project collecting, exchanging and reporting on data from many systems typically has a small number of conceptual models describing the overall scope, more logical models that break that scope into subject areas or functional views, and the most physical models, because each database or system implementation needs its own physical design. The number of models therefore increases from conceptual to logical to physical.

Why the other options are wrong

  • A. Each system implementation adds physical models, so physical models outnumber logical ones.
  • B. Models at different levels do not map one to one when many systems are involved.
  • C. This reverses the usual pattern; conceptual models are the fewest.
  • D. Multiple systems require more than one physical model.

Question 21

In the common enterprise architecture model coded BIAT, the 'I' stands for:

  1. Interoperability
  2. Instance
  3. Identification
  4. Integration
  5. Information
Show answer and explanation

Correct answer: E. Information

The BIAT model divides enterprise architecture into four layers: Business, Information, Application and Technology. Business architecture covers strategy, capabilities and processes; information architecture covers the data and information the business needs and how it is structured; application architecture covers the systems that use that information; and technology architecture covers the infrastructure that hosts them. Information architecture corresponds to the data architecture domain described in DMBOK, so the 'I' stands for Information.

Why the other options are wrong

  • A. Interoperability is a quality of connected systems, not a BIAT layer.
  • B. Instance is not an architecture layer.
  • C. Identification is not an architecture layer.
  • D. Integration is handled within the application and information layers rather than as a layer of its own.

Question 22

Database monitoring tools measure key database metrics, such as:

  1. Capacity, availability, cache performance, user statistics
  2. Capacity, availability, backup instances, data quality
  3. Create, read, normalization, user access
  4. Create, read, update, delete
  5. Capacity, design, normalization, user access
Show answer and explanation

Correct answer: A. Capacity, availability, cache performance, user statistics

DMBOK describes database monitoring tools as automating the monitoring of key metrics such as capacity, availability, cache performance and user statistics. These metrics show whether the database has enough space, whether it is up and reachable, how efficiently memory is used and how users are loading the system. They help DBAs detect problems early and tune performance before users are affected.

Why the other options are wrong

  • B. Data quality is measured with profiling and quality tools, not database monitoring tools.
  • C. Normalization is a design technique, not a runtime metric.
  • D. Create, read, update and delete are data operations, not monitoring metrics.
  • E. Design and normalization are modeling concerns, not monitored database metrics.

Question 23

The creation of overly complex enterprise integration over time is often a symptom of:

  1. Multiple data owners
  2. Multiple integration technologies
  3. Multiple data warehouses
  4. Multiple metadata tags
  5. Multiple application coding languages
Show answer and explanation

Correct answer: B. Multiple integration technologies

When different projects choose their own integration tools and methods over time, the organization ends up with a mix of point-to-point interfaces, ETL tools, messaging and other technologies. Each needs its own skills, monitoring and maintenance, and data flows become hard to trace and change. DMBOK describes this as a typical result of integration built without an enterprise approach, which is why it recommends standard integration patterns, shared services and a limited set of technologies.

Why the other options are wrong

  • A. Multiple data owners are a governance matter and do not in themselves create integration complexity.
  • C. Multiple warehouses may add some flows but are not the typical cause of overly complex enterprise integration.
  • D. Metadata tags describe data and do not create integration complexity.
  • E. Coding languages within applications are not the main source of complexity between systems.

Question 24

Mapping requirements and rules for moving data from source to target enables:

  1. Extract
  2. Backups
  3. Analysis
  4. Load
  5. Transformation
Show answer and explanation

Correct answer: E. Transformation

A mapping defines the sources to be extracted, the targets to be loaded and the rules for converting data from the source structures and values into the target format. DMBOK treats mapping as closely tied to transformation: it is the specification that tells the transformation step how to reformat, convert, derive and validate data as it moves. Without mapping requirements and rules, data cannot be correctly transformed from source to target.

Why the other options are wrong

  • A. Extraction can run without detailed mappings; mappings exist to define how data is converted.
  • B. Backups copy data for recovery and do not need source-to-target rules.
  • C. Analysis of data is a separate activity from moving data between systems.
  • D. Loading writes data that has already been converted according to the mapping rules.

Question 25

A security mechanism that searches for customer bank account details in outgoing emails is achieving the goal of:

  1. Ensuring stakeholder requirements for openness and transparency are met
  2. Ensuring stakeholder requirements for service design and experience are met
  3. Ensuring stakeholder requirements for response time and availability levels are met
  4. Ensuring stakeholder requirements for concise definitions and usage are met
  5. Ensuring stakeholder requirements for confidentiality and privacy are met
Show answer and explanation

Correct answer: E. Ensuring stakeholder requirements for confidentiality and privacy are met

Scanning outgoing email for customer bank account details is a data loss prevention control: it detects sensitive data leaving the organization and can block or flag the message. DMBOK lists among the goals of Data Security ensuring that stakeholder requirements for privacy and confidentiality are met. Customers expect their financial details to stay private, and regulations require organizations to prevent unauthorized disclosure, so this mechanism directly supports confidentiality and privacy.

Why the other options are wrong

  • A. Openness and transparency concern sharing information, whereas this control restricts disclosure.
  • B. Service design and experience are not security goals.
  • C. Response time and availability are performance and operations concerns.
  • D. Concise definitions and usage relate to metadata and business glossaries.

25 questions in. Want the rest? The first 145 questions of the pack, free by email · The full pack

Question 26

Big data is often defined by three characteristics. They are:

  1. Volume, Variety and Velocity
  2. Expansive, Engaged and Enormous
  3. Direction, Depth and Details
  4. Size, Speed and Sensitivity
  5. Complexity, Compliance and Completeness
Show answer and explanation

Correct answer: A. Volume, Variety and Velocity

Big data is commonly defined by the three Vs: volume (the very large quantity of data), variety (structured and unstructured forms such as documents, audio, video and streams) and velocity (the speed at which data is produced and must be processed). DMBOK uses these characteristics in its Big Data and Data Science chapter and notes that further Vs, such as viscosity, volatility and veracity, are sometimes added.

Why the other options are wrong

  • B. These words are not the recognized characteristics of big data.
  • C. Direction, depth and details are not big data characteristics.
  • D. Sensitivity is a security classification, not a defining characteristic of big data.
  • E. Compliance and completeness are governance and quality concerns.

Question 27

An authoritative system where data is created/captured, and/or maintained through a defined set of rules and expectations is called:

  1. A System of Referential Integrity
  2. A System of Retirement
  3. A System of Reference
  4. A System of Systems
  5. A System of Record
Show answer and explanation

Correct answer: E. A System of Record

DMBOK defines a System of Record as an authoritative system where data is created or captured, and/or maintained through a defined set of rules and expectations. An ERP system, for example, may be the System of Record for sell-to customers. It differs from a System of Reference, which provides consumers with trusted data even though that data did not originate there.

Why the other options are wrong

  • A. Referential integrity is a database constraint, not a type of system.
  • B. This is not a recognized system role in data management.
  • C. A System of Reference provides trusted access to data that originated elsewhere.
  • D. A system of systems is a general engineering term for linked systems.

Question 28

A workforce is locked out of a building due to a threat to personnel. The Business Continuity plan (BCP) is stored in the Document Management System (DMS).

Luckily:

  1. The security policy includes such a scenario
  2. The workforce are expendable
  3. The document is printed and stored off site
  4. The metadata repository holds the BCP
  5. The data warehouse catalog is fully operational
Show answer and explanation

Correct answer: C. The document is printed and stored off site

A business continuity plan must be available when the normal environment is not. If staff cannot enter the building, a plan held only in the internal document management system may be unreachable at the moment it is needed. DMBOK advises keeping copies of continuity and recovery plans, with the information needed to act on them, in a secure off-site location. A printed copy stored off site lets the workforce follow the plan without access to the building or its systems.

Why the other options are wrong

  • A. Covering the scenario in a policy does not help if no one can access the plan.
  • B. Treating people as expendable is unacceptable and is not a continuity measure.
  • D. A metadata repository is another internal system that may be unreachable, and it is not where plans are kept.
  • E. A data warehouse catalog does not provide access to the continuity plan.

Question 29

A DMZ is bordered by 2 firewalls. These are between the DMZ and the:

  1. Internet, for added security
  2. Internet and intranet
  3. Internet and internal systems
  4. Korean peninsula
  5. Internet and extranet
Show answer and explanation

Correct answer: B. Internet and intranet

A DMZ (demilitarized zone) is a network segment at the edge of an organization that sits between the untrusted Internet and the organization's own private network, its intranet. It is bordered by two firewalls: an outer one between the DMZ and the Internet, and an inner one between the DMZ and the intranet. Public-facing services such as web servers can sit in the DMZ, while the inner firewall limits what traffic can reach the internal network if a DMZ server is compromised.

Why the other options are wrong

  • A. Only the outer firewall faces the Internet; the inner one separates the DMZ from the intranet.
  • C. The inner firewall protects the organization's private network as a whole, which is the intranet named in the other option.
  • D. This refers to a geographic demilitarized zone, not a network security architecture.
  • E. An extranet is a partner-facing network, not the private network that the inner firewall protects.

Question 30

When trying to integrate a large number of systems, the integration complexities can be reduced by:

  1. The use of a common data model
  2. Using data quality measures and targets
  3. Clear business specification and priorities
  4. The use of SQL
  5. Tackling the largest systems first
Show answer and explanation

Correct answer: A. The use of a common data model

Integrating many systems point to point requires a separate mapping for each pair, so the number of interfaces grows rapidly as systems are added. A common (canonical) data model gives every system a shared, standard format: each system maps once to the common model and exchanges data through it. DMBOK describes this approach in Data Integration and Interoperability as a way to reduce the number of transformations and simplify integration, because each new system needs only one mapping rather than one per partner system.

Why the other options are wrong

  • B. Data quality measures improve the data but do not reduce the number of interfaces and mappings.
  • C. Clear specifications help any project but do not reduce structural integration complexity.
  • D. SQL is a query language and does not reconcile differing data structures across systems.
  • E. Sequencing the work by system size does not reduce the number of mappings required.

Question 31

Basic profiling of data involves analysis of:

  1. Data sources, data visualisations, data latency and ownership
  2. Data lakes, data warehouses and operational data stores
  3. Data criteria, data acquisition, data policy and procedures inside data governance
  4. Data formats, data values, patterns and relationships
  5. Data clusters, data clouds, data repositories and metadata libraries
Show answer and explanation

Correct answer: D. Data formats, data values, patterns and relationships

Data profiling examines actual data to understand its content and structure. Basic profiling analyzes data formats as defined and as found in the data, population levels such as nulls and defaults, data values and how they match the defined valid values, patterns and relationships within the data set, and relationships to other data sets. The results reveal anomalies and help define data quality rules and integration mappings.

Why the other options are wrong

  • A. Visualizations, latency and ownership are not what profiling analyzes.
  • B. These are storage platforms, not the subject of profiling analysis.
  • C. Criteria, acquisition and governance procedures are not profiling activities.
  • E. Clusters, clouds and repositories are infrastructure terms, not profiling analyses.

Question 32

A key feature of Bill Inmon's approach to data warehousing is:

  1. A preference for supporting operational reporting
  2. A normalized relational model to store and manage data
  3. Its ability to operate on open source platforms
  4. An exclusive focus on star schemas and cubes
  5. A tight management of data dimensions
Show answer and explanation

Correct answer: B. A normalized relational model to store and manage data

Bill Inmon defines the data warehouse as a subject-oriented, integrated, time-variant and non-volatile collection of data in support of management's decision making. In his Corporate Information Factory approach, the central enterprise data warehouse uses a normalized relational model to store and manage integrated data, and dimensional data marts are then built from it for specific analysis needs. The normalized relational core distinguishes Inmon's approach from Kimball's dimensional bus architecture.

Why the other options are wrong

  • A. Operational reporting is mainly supported by operational systems and operational data stores, not the defining feature of Inmon's warehouse.
  • C. Inmon's approach is not defined by any platform type.
  • D. An exclusive dimensional focus describes Kimball's approach rather than Inmon's.
  • E. Dimension management is central to Kimball's conformed dimensions, not Inmon's normalized core.

Question 33

The advantage of a decentralized data governance model over a centralized model is:

  1. An increased level of ownership from local decision making groups
  2. The cheaper execution of data governance operations
  3. The easier implementation of industry data models
  4. Having a common approach to resolving data governance issues
  5. The common metadata repository configurations
Show answer and explanation

Correct answer: A. An increased level of ownership from local decision making groups

In a decentralized (federated or distributed) governance model, decision making and stewardship are spread across business units rather than held by one central body. Because local groups make the decisions that affect their own data, they tend to feel stronger ownership of the data and the governance outcomes. The trade-off is less consistency and some duplication of effort, which is why a centralized model is better at providing a common approach to issues.

Why the other options are wrong

  • B. Decentralized governance often costs more because effort is duplicated across units.
  • C. Industry data model adoption does not depend on the governance model.
  • D. A common approach to issues is an advantage of a centralized model.
  • E. Common repository configurations come from centralized coordination.

Question 34

By comparing the system of record and systems of reference to each other, it is possible to:

  1. Validate the accuracy of the master data
  2. Update the core reference values
  3. Construct time variant sequences
  4. Validate the completeness of the master data
  5. Validate the consistency of the master data
Show answer and explanation

Correct answer: E. Validate the consistency of the master data

Comparing the system of record with the systems of reference shows whether the same master data values agree everywhere they are held. Differences reveal synchronization failures, stale copies or local changes that were never propagated, which are consistency problems. This comparison cannot prove accuracy, because two systems can agree on a wrong value, and it cannot prove completeness against business requirements. It is therefore a check of the consistency of the master data across systems.

Why the other options are wrong

  • A. Accuracy is measured against the real world or a verified reference, and matching systems can still hold the same wrong value.
  • B. Comparison detects differences; it does not update core reference values.
  • C. Time-variant sequences come from history tracking, not from comparing systems.
  • D. Completeness is judged against requirements, not by comparing copies of the same data.

Question 35

A relationship that allows an address to be used by multiple people, and each person can have multiple addresses, can be resolved:

  1. With an additional relationship describing the address usage
  2. By changing the primary keys on Person and Address to ensure referential integrity
  3. With an associative entity called Person Address Usage and two, 'one to many' relationships
  4. With a partnership entity called Person Address Usage and two, 'one to many' relationships
  5. By changing the role names of the foreign keys on Person and Address to ensure referential integrity
Show answer and explanation

Correct answer: C. With an associative entity called Person Address Usage and two, 'one to many' relationships

A person can have many addresses and an address can be used by many people, which is a many-to-many relationship. Logical modeling resolves this by adding an associative entity, here Person Address Usage, that holds one instance for each pairing of a person and an address. The original relationship becomes two one-to-many relationships, Person to Person Address Usage and Address to Person Address Usage, and the associative entity can also carry attributes of the pairing, such as the address type or effective dates.

Why the other options are wrong

  • A. Adding another relationship leaves the many-to-many relationship unresolved.
  • B. Changing primary keys does not remove the many-to-many relationship.
  • D. A partnership entity is not a data modeling construct; the construct used is an associative entity.
  • E. Renaming foreign key roles does not resolve the many-to-many relationship.

Question 36

GDPR and PIPEDA are examples of:

  1. Data protection regulations
  2. Primary information parsing algorithms
  3. Global data modelling standards
  4. Data program rules
  5. Content management systems
Show answer and explanation

Correct answer: A. Data protection regulations

The EU General Data Protection Regulation and Canada's Personal Information Protection and Electronic Documents Act are laws governing how organizations collect, use, store and disclose personal data. DMBOK cites both in its discussion of data handling ethics and data security as examples of regulations that set requirements for consent, purpose limitation, individual rights and protection of personal information.

Why the other options are wrong

  • B. They are laws, not parsing algorithms.
  • C. They do not define data modeling standards.
  • D. They are national and regional regulations, not internal program rules.
  • E. They are not software systems.

That was 36 of 295.

The full DAMA CDMP Data Management Fundamentals pack has all 295 questions, each with the answer, the explanation and why the other options are wrong, plus a questions-only copy for timed runs. US$39, paid once, with free monthly updates and a pass-or-your-money-back guarantee.

Get the full pack

Not ready yet? Get the first 145 free as one PDF by email. Get the Half Pack →