Free DAMA CDMP Fundamentals Practice Questions: Big Data

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: Big Data

A retailer is launching an analytics initiative that combines point-of-sale transactions with web clickstream events, product images, customer reviews, and near-real-time inventory sensor feeds. The first release will use less data than the existing enterprise warehouse. Which management approach best fits the initiative?

Options:

• A. Manage it as a traditional warehouse because it is smaller

• B. Prioritize storage expansion before defining governance practices

• C. Require all inputs to use one fixed relational schema first

• D. Plan for varied formats, streaming ingestion, and evolving metadata

Best answer: D

Explanation: Big data concerns are not defined by size alone. In this case, the decisive characteristics are varied data types (transactions, clickstream, images, reviews, sensor feeds), different rates of arrival, and likely schema evolution. Management practices should address scalable ingestion, metadata, lineage, quality, governance, and fit-for-purpose processing across structured, semi-structured, and unstructured data. A smaller first release can still require big data management practices if the data characteristics create complexity beyond traditional structured-data handling.

The key takeaway is to assess the nature and use of the data, not only the number of records or terabytes.

• Smaller than the warehouse fails because size alone does not determine whether big data management concerns apply.
• Storage-first thinking misses governance, metadata, and processing implications created by variety and velocity.
• Fixed relational schema may be useful for some curated outputs, but forcing every source into it upfront can undermine flexible ingestion and analysis.

Question 2

Topic: Big Data

A retailer is moving high-volume clickstream and mobile-app events into a distributed data lake to support near-real-time customer analytics. The platform can scale storage and processing, but event names differ across channels, some fields contain personal data, and business teams want trusted dashboards within the next quarter. What is the best professional decision before expanding data consumption?

Options:

• A. Rely on distributed processing and encryption as sufficient controls

• B. Allow each product team to define events independently

• C. Prioritize ingestion speed and defer controls until reporting stabilizes

• D. Define governance, quality, metadata, and security controls for the data lake

Best answer: D

Explanation: Big data platforms change scale, velocity, and variety, but they do not remove core data-management responsibilities. In this scenario, inconsistent event names create a quality and semantic issue, personal data creates security and privacy obligations, and trusted dashboards require shared definitions and metadata. A professional response is to establish practical controls such as accountable ownership, common event definitions, data quality rules, cataloging, lineage, classification, and access management. These controls can be lightweight and iterative, but they must be present before broad consumption. Platform scalability helps process data; it does not make the data understandable, trustworthy, or appropriately protected.

• Deferring controls increases the risk that dashboards are built on inconsistent, poorly understood event data.
• Local event definitions may speed delivery but worsen semantic inconsistency across channels.
• Encryption alone protects data confidentiality in limited ways but does not address quality, metadata, lineage, or stewardship.

Question 3

Topic: Big Data

A logistics company is adding GPS telemetry from delivery vehicles to its data platform. Events arrive continuously and must support route-exception alerts within 10 seconds. The current nightly batch load cannot meet the requirement. Which big data characteristic most directly drives the needed change in ingestion and processing design?

Options:

• A. Veracity

• B. Variety

• C. Volume

• D. Velocity

Best answer: D

Explanation: Velocity describes how quickly data is generated, ingested, and must be acted upon. In this case, GPS events arrive continuously and must trigger route-exception alerts within 10 seconds, so the platform needs streaming or near-real-time ingestion and processing rather than only a nightly batch load. Volume might influence storage scaling, variety might influence schema and format handling, and veracity might influence trust and quality controls, but the visible constraint is time sensitivity. The key takeaway is to match the big data characteristic to the management decision it most directly affects.

• Volume would matter most if the main constraint were data size, storage capacity, or large-scale retention.
• Variety would matter most if the main issue were multiple formats, structures, or semantic differences across sources.
• Veracity would matter most if the main concern were uncertainty, accuracy, bias, or trustworthiness of the telemetry.

Question 4

Topic: Big Data

A utility receives semi-structured JSON events from thousands of field sensors. Events are used for near-real-time outage alerts, but the same raw events must also be retained for later analysis and reprocessing. Some devices intermittently send impossible voltage values and duplicate event IDs. Which control point best fits the data lifecycle need?

Options:

• A. Merge sensor events into customer master data

• B. Correct values only in the BI dashboard layer

• C. Delete raw events immediately after alert generation

• D. Validate and quarantine exceptions during stream ingestion

Best answer: D

Explanation: Big data lifecycle controls should be placed where they protect the earliest dependent use without destroying future value. For streaming sensor data used in near-real-time decisions, validation, de-duplication checks, tagging, and exception quarantine should occur during ingestion or stream processing before alerts are published. Retaining raw events separately supports replay, audit, data science, and root-cause analysis. This approach recognizes big data implications: high velocity, semi-structured formats, and late-arriving or imperfect data. Fixing only downstream reports is too late for alerting, and deleting raw data removes evidence needed for improvement and reprocessing.

• Dashboard-only correction fails because alerts have already been triggered before BI presentation controls apply.
• Master data merge confuses sensor event data with stable core business entities such as customers or assets.
• Immediate deletion undermines replay, auditability, and later analytical use of raw big data.

Question 5

Topic: Big Data

A retail company receives millions of clickstream and mobile-app events per hour. Marketing wants near-real-time personalization within seconds, while compliance requires traceable consent status, retention rules, and the ability to replay events after processing failures. Which management approach best fits these requirements?

Options:

• A. Manual stewardship review before each event is processed

• B. Nightly batch loads into the enterprise warehouse

• C. Governed streaming ingestion with replayable event logs

• D. Uncontrolled raw data lake for all events

Best answer: C

Explanation: High-velocity data often requires a streaming management approach rather than only periodic batch processing. In this case, the business value depends on seconds-level personalization, so events must be ingested and processed continuously. Reliability and control still matter: replayable event logs help recover from failures, while governance practices such as metadata capture, consent tracking, retention rules, and lineage make the data usable and accountable. Big data management is not just about storing large volumes; it must align processing patterns with business latency needs and control obligations.

A raw repository alone may handle volume, but it does not provide the governed, reliable event handling required here.

• Nightly batch loading misses the seconds-level latency requirement, even though it may support historical reporting.
• Raw data lake only may store volume cheaply but lacks the stated controls for consent, retention, lineage, and recovery.
• Manual event review creates an operational bottleneck and is not feasible for millions of events per hour.

Question 6

Topic: Big Data

A company is moving high-volume customer interaction data from a single relational warehouse into a distributed data lake and scalable processing platform. Business users expect faster analytics, but the data will still support regulatory reports and customer segmentation. Which statement best distinguishes the platform capability from the continuing data-management responsibilities?

Options:

• A. Raw data zones should be exempt from ownership and access controls.

• B. Scalability increases processing capacity, while governance, quality, metadata, and security controls remain necessary.

• C. Schema-on-read removes the need for business definitions and lineage.

• D. Replication across nodes makes formal data quality rules unnecessary.

Best answer: B

Explanation: Big data technologies address characteristics such as volume, velocity, and scalability, but they do not remove core data-management obligations. Data still needs accountable ownership, stewardship, quality expectations, metadata for meaning and lineage, and security controls based on sensitivity and allowed use. A distributed platform may change how controls are implemented, but the need for trusted, well-understood, and protected data remains. This is especially important when outputs support regulated reporting or customer analytics.

• Replication misconception fails because multiple copies can improve availability, not correctness or fitness for use.
• Schema-on-read misconception fails because delayed physical structure does not eliminate business meaning, definitions, or lineage.
• Raw-zone exemption fails because sensitive data and accountability requirements apply before data is curated.

Question 7

Topic: Big Data

A utility company receives streaming sensor readings and gateway logs from thousands of field devices into a data lake. Analysts report that some events have missing device identifiers or timestamps outside the operating period. The team must protect downstream analytics without losing the original raw events. Which lifecycle control point best fits this need?

Options:

• A. Redesign the warehouse star schema

• B. Archive raw streams before profiling

• C. Correct values only in BI reports

• D. Validate and quarantine exceptions during ingestion

Best answer: D

Explanation: Big data lifecycle thinking places controls where they reduce risk without destroying useful raw data. For streaming sensor and log data, the ingestion or landing stage is an appropriate control point for basic validation, metadata capture, and exception handling. Records with missing identifiers or impossible timestamps can be flagged or quarantined while the raw event stream remains available for audit, replay, or later correction. This supports timely analytics and avoids pushing known defects into curated zones, warehouses, or BI products. Later modelling or reporting layers may consume the cleansed or certified data, but they should not be the first place obvious ingestion defects are discovered or handled.

• BI-only correction hides defects late in the lifecycle and can leave inconsistent results across other downstream uses.
• Warehouse redesign addresses analytical structure, not early validation of streaming event defects.
• Premature archiving delays profiling and quality checks, allowing bad data to remain unmanaged before use.

Question 8

Topic: Big Data

A retailer is launching a customer insight platform that will ingest point-of-sale transactions, clickstream events, product reviews, and social media comments. Events arrive continuously, formats differ by source, daily volumes are growing quickly, and marketing wants trusted customer sentiment metrics for campaign decisions. Which data management approach best fits these big data characteristics?

Options:

• A. Use scalable ingestion with metadata, quality rules, and governance controls

• B. Standardize only the transaction schema before storing any data

• C. Load all sources into a single spreadsheet for analyst review

• D. Postpone governance until the platform reaches stable production use

Best answer: A

Explanation: Big data characteristics shape management choices. High volume and velocity call for scalable ingestion, storage, and processing patterns. Variety requires metadata, schema management, and integration practices that can handle structured transactions, semi-structured clickstream data, and unstructured text. Veracity requires quality controls, source understanding, lineage, and stewardship so analytical outputs can be trusted. Governance should be designed into the platform early because definitions, access, retention, quality expectations, and accountability affect how the data can be used for business decisions.

• Spreadsheet consolidation cannot handle continuous arrival, growing volume, or controlled reuse across diverse sources.
• Transaction-only standardization ignores clickstream, reviews, and social content, so it does not address variety or sentiment needs.
• Delayed governance increases risk because trusted metrics depend on definitions, lineage, quality rules, and accountability from the start.

Question 9

Topic: Big Data

A data management team is classifying new initiatives. They agree that dataset size alone should not determine whether big data management practices are needed. Which initiative most clearly raises big data management concerns rather than traditional structured-data management concerns?

Options:

• A. Ten-year ERP history loaded monthly to a warehouse

• B. Very large relational customer table with stable columns

• C. Continuous clickstream, sensor, and log feeds with evolving formats

• D. Nightly reference-code synchronization across applications

Best answer: C

Explanation: Big data management is not defined by size alone. DAMA-DMBOK frames big data concerns around characteristics such as volume, velocity, variety, variability, and the need for scalable processing, governance, metadata, quality, and security approaches across diverse data forms. A very large but stable relational table can often be managed with traditional structured-data practices. Continuous feeds from clickstream, sensors, and logs introduce varied formats, rapid arrival, changing structures, and different ingestion and interpretation needs. Those characteristics make the management approach different, not merely the number of bytes stored.

• Large table size is tempting, but stable relational columns can still fit traditional structured-data management.
• Monthly ERP history adds historical volume, but the source and load pattern remain conventional and structured.
• Reference-code synchronization is mainly a reference data and integration concern, not a big data discriminator.

Question 10

Topic: Big Data

A retailer wants to use clickstream events, mobile-app telemetry, product reviews, and loyalty-customer records to improve churn prediction. The data arrives continuously, includes semi-structured and unstructured content, and has uneven consent and lineage information. Storage volume is high but manageable on current platforms. Which professional decision best reflects the big data management concern?

Options:

• A. Classify it as big data because the total storage footprint is large

• B. Govern for variety, velocity, provenance, privacy, and fit-for-purpose quality

• C. Force all sources into the existing relational warehouse model first

• D. Allow unrestricted raw-data access until the model proves value

Best answer: B

Explanation: Big data management is driven by characteristics such as variety, velocity, variability, provenance, and intended analytical use, not by size alone. In this scenario, the retailer must manage streaming behavior data, semi-structured telemetry, unstructured reviews, and governed customer master data together. The uneven consent and lineage information also creates privacy, metadata, and trust concerns before broad analytical use. A sound professional response is to apply governance, metadata, integration, privacy, and fit-for-purpose quality controls appropriate to these characteristics, while still supporting exploration and analytics. Treating volume as the only trigger misses the main management implications.

• Volume-only reasoning fails because the current platforms can handle the storage, and the harder issues are variety, velocity, provenance, and consent.
• Warehouse-first modelling is too rigid for streaming, semi-structured, and unstructured sources before their analytical value and meaning are understood.
• Unrestricted raw access ignores privacy, lineage, and customer-data governance risks in the stated facts.

Continue in the web app

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

Try DAMA CDMP Data Management Fundamentals on Web

Related focused pages

• Free DAMA CDMP Data Management Fundamentals Full-Length Practice Exam
• Data Management Process
• Data Architecture
• Document and Content Management
• Data Ethics
• Data Governance
• Data Integration and Interoperability
• Master and Reference Data Management
• Data Modelling and Design
• Data Quality
• Data Security
• Data Storage and Operations
• Data Warehousing and Business Intelligence
• Metadata Management