I85PREP
⌘ K
DMLI study materials
DMLI study resources

DMLI practice questions

One question at a time. Build your DMLI understanding.

First 50 of 200 questions. Each one lists the choices, the correct answer, and an explanation. The checker at the top of this page saves progress in this browser when it is available.

Continue in the practice tool above to load more questions, 50 at a time.

Question 1Data Management Fundamentals

In relational theory, what does a primary key uniquely do for a table?

  1. Identifies each row Correct answer.
  2. Duplicates each row
  3. Sorts each row
  4. Encrypts each row

Explanation

A primary key is a column or set of columns whose values uniquely identify every row in a relation and cannot be null. It enforces entity integrity. It does not impose ordering, encryption, or duplication.

Question 2Data Management Fundamentals

Which property of a database transaction ensures that either all of its operations are applied or none are?

  1. Durability
  2. Atomicity Correct answer.
  3. Isolation
  4. Consistency

Explanation

Atomicity treats a transaction as a single all-or-nothing unit, so partial changes are never committed. Durability concerns persistence after commit, isolation concerns concurrent visibility, and consistency concerns valid state transitions.

Question 3Data Management Fundamentals

What is the main purpose of normalizing a relational schema?

  1. Reduce redundancy and update anomalies Correct answer.
  2. Eliminate the need for keys
  3. Increase storage redundancy
  4. Speed up every query

Explanation

Normalization organizes attributes into relations to remove redundancy and the insertion, update, and deletion anomalies it causes. It does not remove the need for keys and can sometimes add joins that affect performance.

Question 4Data Management Fundamentals

Which of the following best describes metadata?

  1. A physical storage block
  2. A backup copy of a table
  3. Data that describes other data Correct answer.
  4. Raw transactional records

Explanation

Metadata is structured information describing data, such as definitions, formats, ownership, and lineage. Raw records are the data itself, and backups and storage blocks are physical artifacts rather than descriptive information.

Question 5Data Management Fundamentals

A foreign key in a child table references the primary key of a parent table. What does the foreign key constraint enforce?

  1. Uniqueness of child rows
  2. Referential integrity between the tables Correct answer.
  3. Sorting of child rows
  4. Compression of parent rows

Explanation

A foreign key constraint ensures each non-null child value matches an existing parent key value, preserving referential integrity. It does not require child values to be unique or impose ordering or storage effects.

Question 6Data Management Fundamentals

What distinguishes structured data from unstructured data?

  1. Structured data is always encrypted
  2. Structured data fits a predefined model Correct answer.
  3. Structured data lacks any schema
  4. Structured data cannot be queried

Explanation

Structured data conforms to a predefined schema such as relational tables, making it directly queryable. Unstructured data lacks such a model, but structured data is neither inherently encrypted nor unqueryable.

Question 7Data Management Fundamentals

In data governance, what is a data steward primarily responsible for?

  1. Software license purchasing
  2. Physical server maintenance
  3. Network firewall configuration
  4. Day-to-day quality and definition of data Correct answer.

Explanation

A data steward manages the operational care of a data domain, including definitions, quality, and issue resolution. Server maintenance, firewall configuration, and license purchasing are IT infrastructure or procurement duties.

Question 8Data Management Fundamentals

What is the role of a data dictionary?

  1. Storing only numeric data
  2. Replacing the database engine
  3. Encrypting stored records
  4. Documenting definitions of data elements Correct answer.

Explanation

A data dictionary documents data elements, their meanings, types, and constraints, supporting consistency and governance. It is documentation rather than a storage engine, encryption tool, or type restriction.

Question 9Data Management Fundamentals

A table stores customer city in the same row as each order. An update changes the city for one order but not others, producing conflicting values. Which design problem does this illustrate?

  1. A deadlock from ordering
  2. A lost update from locking
  3. A phantom read from isolation
  4. An update anomaly from redundancy Correct answer.

Explanation

Storing the same fact in many rows creates redundancy, so an update applied inconsistently yields an update anomaly. Locking, isolation, and deadlocks are concurrency phenomena, not schema redundancy problems.

Question 10Data Management Fundamentals

A relation has candidate keys A and B. Which statement is true about primary keys?

  1. Primary keys may contain nulls
  2. The primary key is one chosen candidate key Correct answer.
  3. Every candidate key must be primary
  4. Candidate keys must be composite

Explanation

All candidate keys are minimal unique identifiers, and the designer selects one as the primary key. The others remain alternate keys, and primary keys cannot contain nulls. Candidate keys may be single or composite.

Question 11Data Management Fundamentals

A transaction reads a row, another transaction updates and commits that row, and the first transaction re-reads and sees a different value. Which phenomenon occurred?

  1. Non-repeatable read Correct answer.
  2. Phantom read
  3. Dirty read
  4. Lost update

Explanation

A non-repeatable read occurs when re-reading the same row within a transaction returns a changed committed value. A dirty read would involve uncommitted data, a phantom involves new rows matching a predicate, and a lost update involves overwriting another's change.

Question 12Data Management Fundamentals

Which normal form specifically addresses removing partial dependencies of non-key attributes on part of a composite key?

  1. First normal form
  2. Boyce-Codd normal form
  3. Second normal form Correct answer.
  4. Third normal form

Explanation

Second normal form requires a relation to be in first normal form and to have no partial dependency of a non-prime attribute on part of a candidate key. First normal form addresses atomic values, while third and Boyce-Codd forms address transitive and other determinant dependencies.

Question 13Data Management Fundamentals

A column should store monetary amounts with exact precision. Which data type choice is most appropriate?

  1. A fixed-precision decimal type Correct answer.
  2. A boolean type
  3. A single-precision float
  4. A variable-length text type

Explanation

Fixed-precision decimal types represent exact values, avoiding the rounding error inherent in binary floating point for money. Text and boolean types cannot represent amounts with arithmetic precision.

Question 14Data Management Fundamentals

What is the key difference between a data lake and a data warehouse?

  1. A lake cannot store files; a warehouse cannot store tables
  2. A lake requires a rigid schema; a warehouse stores only unstructured files
  3. A lake stores raw varied data; a warehouse stores modeled structured data Correct answer.
  4. A lake and a warehouse differ only in vendor

Explanation

A data lake typically holds raw data in native formats with schema applied on read, while a data warehouse stores curated, modeled structured data optimized for analytics. The distinction is architectural, not vendor-specific.

Question 15Data Management Fundamentals

A data modeler needs to represent many-to-many relationships between students and courses in a relational database. What is the standard approach?

  1. Duplicate the student table for each course
  2. Use a boolean column on the course table
  3. Store a list of courses in a single student column
  4. Create an associative table with foreign keys to both entities Correct answer.

Explanation

Relational tables cannot directly hold many-to-many links, so an associative (junction) table with foreign keys to both entities resolves it into two one-to-many relationships. Multi-valued columns violate first normal form, and duplicating or boolean columns cannot represent the pairing.

Question 16Data Management Fundamentals

Which practice best supports data lineage tracking?

  1. Avoiding documentation of transformations
  2. Deleting intermediate staging tables
  3. Storing data only in flat files
  4. Recording source-to-target mappings for each transformation Correct answer.

Explanation

Lineage is built by capturing how data moves and transforms from source to target, including mappings and transformation logic. Removing staging tables or documentation and using only flat files erodes the ability to trace data origins.

Question 17Data Management Fundamentals

An organization needs a single trusted definition of 'active customer' used across departments. Which governance artifact best provides this?

  1. A business glossary entry Correct answer.
  2. A firewall rule
  3. A storage tier policy
  4. A backup schedule

Explanation

A business glossary captures agreed business definitions and terms so departments share one meaning. Firewall rules, backup schedules, and tiering policies are operational controls that do not define business terms.

Question 18Data Management Fundamentals

What does the consistency property of ACID primarily guarantee?

  1. Committed data survives power loss
  2. Transactions move the database between valid states Correct answer.
  3. Transactions run in parallel without locks
  4. Reads never see the same row twice

Explanation

Consistency requires that a transaction take the database from one valid state to another, respecting declared constraints and rules. Parallelism, crash survival, and read stability are governed by isolation and durability, not consistency.

Question 19Data Management Fundamentals

A column contains the value '2023-13-45' for a date. Which data quality dimension is most directly violated?

  1. Completeness
  2. Validity Correct answer.
  3. Uniqueness
  4. Timeliness

Explanation

Validity concerns conformance to defined formats and domains, and an impossible calendar date fails that rule. Uniqueness concerns duplicates, timeliness concerns currency, and completeness concerns missing values.

Question 20Data Management Fundamentals

A relation is in third normal form. Which dependency has been removed relative to second normal form?

  1. Any dependency on a candidate key
  2. Partial dependency on a composite key
  3. Transitive dependency of non-key attributes Correct answer.
  4. Multi-valued atomicity violation

Explanation

Third normal form removes transitive dependencies where a non-prime attribute depends on another non-prime attribute. Partial dependencies are removed by second normal form, and atomicity is a first normal form concern.

Question 21Data Management Fundamentals

A team must choose between a star schema and a highly normalized model for an analytical warehouse. Which trade-off is most accurate?

  1. Normalized models always query faster for analytics
  2. Normalized models cannot enforce integrity
  3. Star schemas eliminate all data redundancy
  4. Star schemas favor simpler analytical queries at some redundancy cost Correct answer.

Explanation

Star schemas denormalize dimensions to simplify and speed analytical queries, accepting controlled redundancy. Normalized models reduce redundancy and enforce integrity well but often require more joins for analytics.

Question 22Data Management Fundamentals

A relation R(A,B,C) has functional dependencies A to B and B to C, with A the only candidate key. Which normal form does R fail and why?

  1. Second normal form, due to a partial dependency on A
  2. Third normal form, due to a transitive dependency A to B to C Correct answer.
  3. First normal form, due to non-atomic values
  4. Boyce-Codd normal form only, due to a multi-valued dependency

Explanation

Because B is not a candidate key and C depends on B, C is transitively dependent on key A, violating third normal form. There is no composite key, so partial dependency does not apply, and no multi-valued dependency is stated.

Question 23Data Management Fundamentals

Under the read committed isolation level, which anomaly is still possible?

  1. Non-repeatable read Correct answer.
  2. Seeing partially applied transactions
  3. Dirty read
  4. Reading uncommitted writes

Explanation

Read committed prevents dirty reads by exposing only committed data, but a row re-read later can reflect another transaction's commit, producing a non-repeatable read. Dirty reads and uncommitted visibility are already excluded at this level.

Question 24Data Management Fundamentals

A dimension table uses a surrogate key instead of the natural business key. What is the primary benefit in a warehouse?

  1. It removes the need for any keys
  2. It guarantees the dimension is normalized
  3. It eliminates the need for fact tables
  4. It insulates history from changes in source business keys Correct answer.

Explanation

Surrogate keys decouple warehouse rows from volatile source identifiers, allowing multiple versions of a member to be tracked without altering fact references. They do not remove keys, force normalization, or replace fact tables.

Question 25Data Management Fundamentals

An analyst must reconcile conflicting values for the same customer across two source systems. Which governance mechanism most directly resolves which value is authoritative?

  1. A faster network link
  2. A daily full backup
  3. A data ownership model naming the system of record Correct answer.
  4. A larger storage tier

Explanation

Assigning data ownership and designating a system of record establishes which source is authoritative when values conflict. Storage, network, and backup choices affect operations but do not adjudicate data authority.

Question 26Data Management Fundamentals

In a relational database, what is the primary purpose of a foreign key?

  1. To uniquely identify each row within its own table
  2. To automatically generate sequential numeric values
  3. To encrypt sensitive columns at rest
  4. To enforce referential integrity between two tables Correct answer.

Explanation

A foreign key is a column or set of columns whose values must match a candidate key in another (or the same) table, which enforces referential integrity. Uniqueness within a table is the job of a primary or unique key, not a foreign key.

Question 27Data Management Fundamentals

Which characteristic best describes a relational database's handling of a NULL value in a column?

  1. It always causes the row to be rejected by the database
  2. It is treated as an empty string in every data type
  3. It represents an unknown or absent value rather than zero or blank Correct answer.
  4. It is automatically replaced by the column default on insert

Explanation

In relational theory NULL denotes the absence of a value, distinct from zero, empty string, or any concrete value. Comparisons with NULL therefore use three-valued logic (IS NULL) rather than equality.

Question 28Data Management Fundamentals

A data architect needs to record the technical structure of tables, columns, data types, and allowed values for a new system. Which artifact is designed for this purpose?

  1. A data dictionary Correct answer.
  2. A business glossary
  3. A data flow diagram
  4. A retention schedule

Explanation

A data dictionary documents technical metadata such as column names, data types, lengths, and permitted values. It differs from a business glossary, which defines business terms rather than physical structures.

Question 29Data Management Fundamentals

Which statement correctly describes the difference between a primary key and a unique constraint in most relational databases?

  1. A primary key can span multiple tables, while a unique constraint cannot
  2. A primary key enforces referential integrity, while a unique constraint enforces domain integrity
  3. A primary key disallows NULLs, while a unique constraint may permit them Correct answer.
  4. A primary key may be duplicated, while a unique constraint may not

Explanation

A primary key both enforces uniqueness and requires non-null values, whereas a unique constraint enforces uniqueness but generally allows NULLs. Each table can have only one primary key but multiple unique constraints.

Question 30Data Management Fundamentals

A relation stores order data with columns OrderID, ProductID, Quantity, and ProductName, where ProductName depends only on ProductID. The key is the combination (OrderID, ProductID). Which normalization issue does this represent?

  1. A transitive dependency on a non-key attribute
  2. A multivalued dependency
  3. A join dependency
  4. A partial dependency on part of a composite key Correct answer.

Explanation

ProductName depends on only ProductID, which is part of the composite key, so it is a partial dependency that violates second normal form. Transitive dependencies involve non-key attributes determining other non-key attributes, and multivalued or join dependencies relate to higher normal forms.

Question 31Data Management Fundamentals

A customer support system must record that each ticket may be assigned to several agents and each agent may handle several tickets. Which schema design correctly models this in a relational database?

  1. A foreign key from agents to tickets only
  2. A repeating group of agent columns in the tickets table
  3. A single foreign key from tickets to agents
  4. A junction table with foreign keys to both tickets and agents Correct answer.

Explanation

Many-to-many relationships are resolved with an associative (junction) table whose composite key references both entity tables. This avoids repeating groups and preserves referential integrity.

Question 32Data Management Fundamentals

A transaction transfers funds between two accounts. Midway through, the system crashes before commit. Which ACID property ensures the partial transfer is undone?

  1. Consistency
  2. Atomicity Correct answer.
  3. Isolation
  4. Durability

Explanation

Atomicity guarantees that a transaction is treated as a single all-or-nothing unit, so an incomplete transaction is rolled back. Durability concerns committed changes surviving failure, and isolation concerns concurrent transaction visibility.

Question 33Data Management Fundamentals

An analyst queries a table and finds that some rows have a birth date later than the current date. Which data quality dimension is most directly violated?

  1. Uniqueness
  2. Timeliness
  3. Completeness
  4. Validity Correct answer.

Explanation

Validity concerns whether values conform to defined rules and permissible ranges, such as a birth date not exceeding today. Completeness concerns missing values, uniqueness concerns duplicates, and timeliness concerns data currency.

Question 34Data Management Fundamentals

A company wants to assign a stable, system-generated integer key to each row in a dimension table so that source system changes do not break existing references. What is this key called?

  1. A surrogate key Correct answer.
  2. A natural key
  3. A composite key
  4. A foreign key

Explanation

A surrogate key is a system-assigned identifier with no business meaning, used to insulate the warehouse from changes in source business keys. A natural key is derived from real-world business data and may change.

Question 35Data Management Fundamentals

An organization wants to track where each data element originated, how it was transformed, and where it is consumed. Which capability does this describe?

  1. Data masking
  2. Data replication
  3. Data lineage Correct answer.
  4. Data archiving

Explanation

Data lineage records the origin, movement, and transformation history of data across systems. Masking, replication, and archiving address protection, copying, and retention respectively.

Question 36Data Management Fundamentals

A data engineer must store a column that holds free-form customer comments of varying length in a relational database. Which design choice is most appropriate?

  1. A fixed-length character type sized to the realistic maximum
  2. A numeric type with a large precision
  3. A date type with a text conversion
  4. A variable-length character type sized to the realistic maximum Correct answer.

Explanation

Variable-length character types store only the actual characters plus overhead, which suits free-form text of varying length. Fixed-length types pad every value to the declared size, wasting space.

Question 37Data Management Fundamentals

A relation R(A, B, C, D) has functional dependencies A to B, B to C, and C to D, with A the only candidate key. Which normal form does R currently satisfy at minimum, and which does it violate?

  1. It is in BCNF and violates no normal form
  2. It is in 3NF but violates BCNF because of a partial dependency
  3. It is in 2NF but violates 3NF because of transitive dependencies Correct answer.
  4. It is in 1NF but violates 2NF because of a partial dependency

Explanation

Because A is the sole key, there are no partial dependencies, so R is in 2NF. However A determines B, B determines C, and C determines D, so non-key attributes determine other non-key attributes, violating 3NF.

Question 38Data Management Fundamentals

A team is choosing between a star schema and a third normal form model for a reporting warehouse. Which trade-off best describes the decision?

  1. The star schema simplifies queries at the cost of some redundant data Correct answer.
  2. The star schema eliminates all redundancy at the cost of complex joins
  3. The normalized model speeds up aggregation at the cost of storage
  4. The normalized model is preferred because it removes the need for surrogate keys

Explanation

A star schema denormalizes dimension data, reducing join complexity for analysts while accepting some redundancy. A normalized model reduces redundancy but requires more joins, which can slow analytical queries.

Question 39Data Management Fundamentals

Two transactions run concurrently. Transaction T1 reads a set of rows matching a condition, T2 inserts a new row that satisfies that condition and commits, and T1 re-runs the same query and sees the new row. Which anomaly has occurred?

  1. A lost update
  2. A phantom read Correct answer.
  3. A non-repeatable read
  4. A dirty read

Explanation

A phantom read occurs when re-executing a range or predicate query returns new rows inserted by another committed transaction. A non-repeatable read concerns a changed value in an already-read row, not a new row.

Question 40Data Management Fundamentals

A data steward must classify each attribute in a customer table according to whether it identifies the entity, describes the entity, or records a relationship to another entity. Which foundational data modeling concept is being applied?

  1. Encrypting sensitive attributes so that only authorized staff can view them
  2. Assigning attributes to entity types by role within the entity-relationship model Correct answer.
  3. Versioning each row so that historical changes to customer details are retained
  4. Normalizing the table until every non-key attribute is fully dependent on the whole key

Explanation

Entity-relationship modeling classifies attributes by their role: identifier attributes distinguish entity instances, descriptive attributes capture properties, and relationship attributes (or foreign keys) link entities. This classification is a modeling activity independent of normalization form, versioning strategy, or security controls.

Question 41Data Governance and Compliance

Which principle requires an organization to collect only the personal data that is necessary for a stated purpose?

  1. Data minimization Correct answer.
  2. Accuracy
  3. Purpose limitation
  4. Storage limitation

Explanation

Data minimization limits collection and processing to what is adequate, relevant, and necessary for the stated purpose. Purpose limitation restricts use to the original purpose, while storage limitation governs retention periods.

Question 42Data Governance and Compliance

A retail company wants to use customer email addresses collected for order delivery to send marketing newsletters. Which principle is most directly implicated?

  1. Integrity and confidentiality
  2. Data minimization
  3. Purpose limitation Correct answer.
  4. Accountability

Explanation

Purpose limitation requires that data collected for one purpose not be reused for an incompatible purpose without a lawful basis. Reusing delivery data for marketing is a new purpose requiring separate justification.

Question 43Data Governance and Compliance

Under a typical data protection regime, what is the primary obligation of a data controller?

  1. Auditing the controller's compliance on behalf of a regulator
  2. Processing data strictly on documented instructions of another party
  3. Providing cloud infrastructure for storing personal data
  4. Determining the purposes and means of processing personal data Correct answer.

Explanation

A controller determines why and how personal data is processed and bears the primary compliance responsibility. A processor acts on the controller's documented instructions instead.

Question 44Data Governance and Compliance

A company hires a payroll vendor that processes employee data only on the company's documented instructions. What role does the vendor hold?

  1. Data processor Correct answer.
  2. Data subject
  3. Supervisory authority
  4. Data controller

Explanation

A processor handles personal data on behalf of and on the documented instructions of a controller. The hiring company remains the controller for the payroll data.

Question 45Data Governance and Compliance

Which right allows an individual to obtain a copy of the personal data an organization holds about them?

  1. Right to rectification
  2. Right to erasure
  3. Right of access Correct answer.
  4. Right to restriction of processing

Explanation

The right of access entitles individuals to confirmation of processing and a copy of their personal data. Rectification corrects inaccurate data, and erasure removes it.

Question 46Data Governance and Compliance

An individual exercises the right to erasure. In which situation may an organization lawfully decline the request?

  1. The data was collected more than one year ago
  2. The organization finds the request inconvenient to process
  3. The data is needed to comply with a legal retention obligation Correct answer.
  4. The individual has not explained the reason for the request

Explanation

Erasure rights are subject to exceptions, including compliance with a legal obligation or establishment of legal claims. The other listed reasons are not valid grounds to refuse.

Question 47Data Governance and Compliance

What does data portability primarily require an organization to do?

  1. Transfer all data to a government registry
  2. Delete data on request within thirty days
  3. Provide an individual's data in a structured, commonly used, machine-readable format Correct answer.
  4. Encrypt all personal data at rest

Explanation

Portability lets individuals receive their data in a structured, commonly used, machine-readable format and transmit it to another controller. It is distinct from deletion or security duties.

Question 48Data Governance and Compliance

Which lawful basis typically requires a freely given, specific, informed, and unambiguous indication of the individual's wishes?

  1. Consent Correct answer.
  2. Performance of a contract
  3. Legitimate interest
  4. Legal obligation

Explanation

Valid consent must be freely given, specific, informed, and unambiguous, expressed by a clear affirmative action. The other bases are grounded in contractual, legal, or interest-based necessity rather than the individual's wishes.

Question 49Data Governance and Compliance

A hospital processes patient data to provide emergency medical care. Which lawful basis is most appropriate?

  1. Vital interests Correct answer.
  2. Contract
  3. Consent
  4. Legitimate interest

Explanation

Vital interests apply where processing is necessary to protect someone's life, such as emergency care where consent cannot be obtained. Consent would often be impossible in such circumstances.

Question 50Data Governance and Compliance

Under common breach notification rules, within what general timeframe must a controller notify the supervisory authority of a qualifying personal data breach?

  1. Within 72 hours of becoming aware Correct answer.
  2. Within 30 days of becoming aware
  3. Within 24 hours of becoming aware
  4. At the next annual filing

Explanation

Many regimes require notification to the supervisory authority without undue delay and, where feasible, within 72 hours of awareness. Longer or event-driven timelines do not meet this standard.