Data integrity

Data integrity is the maintenance of, and the assurance of, data accuracy and consistency over its entire life-cycle. It is a critical aspect of the design, implementation, and usage of any system which stores, processes, or retrieves data. The term is broad in scope and may have widely different meanings depending on the specific context – even under the same general umbrella of computing. It is often contrasted with data security, which is the protection of data from unauthorized parties.

There are several types of data integrity: entity integrity, referential integrity, domain integrity, and user-defined integrity. Entity integrity ensures that each row in a table is uniquely identified by a primary key. Referential integrity ensures that relationships between tables remain consistent, typically enforced through foreign keys. Domain integrity ensures that data in a column falls within a defined set of valid values. User-defined integrity refers to rules defined by users to meet specific business requirements.¹

Data integrity is enforced through a variety of mechanisms, including database constraints, transaction processing, and error detection algorithms. In relational databases, constraints such as NOT NULL, UNIQUE, CHECK, and PRIMARY KEY enforce integrity at the schema level. Transactions, governed by the ACID properties (Atomicity, Consistency, Isolation, Durability), ensure that database operations are processed reliably and that integrity is maintained even in the event of failures.²

Beyond databases, data integrity is ensured in data transmission and storage through checksums, hash functions, and error-correcting codes. For example, the MD5 and SHA-1 algorithms are used to verify file integrity, though they are now considered weak for security purposes. In distributed systems, techniques like consensus algorithms (e.g., Paxos, Raft) help maintain consistency and integrity across replicas.³

Data integrity faces numerous threats, including human error, hardware failures, software bugs, and malicious attacks. Human error is the most common cause, such as accidental deletion or incorrect data entry. Hardware failures, like disk crashes, can corrupt data. Software bugs may introduce logical errors that compromise integrity. Malicious attacks, such as SQL injection or ransomware, can deliberately alter or destroy data.

In the era of big data and cloud computing, maintaining integrity becomes more complex due to the scale and distribution of data. Techniques like data provenance, which tracks the origin and history of data, are increasingly important. Additionally, blockchain technology offers a decentralized approach to ensuring data integrity through cryptographic hashing and consensus mechanisms.⁴

Beyond the standard definitions, data integrity has subtle dimensions. For instance, in scientific research, data integrity is a cornerstone of reproducibility; the Open Science movement emphasizes transparent data management to ensure integrity. In regulatory contexts, such as the FDA's 21 CFR Part 11, data integrity is a legal requirement for electronic records in clinical trials.

Another niche area is data integrity in embedded systems and IoT devices, where resource constraints require lightweight integrity checks. Also, the concept of data integrity extends to data quality, which includes accuracy, completeness, and timeliness, but integrity specifically focuses on the absence of unauthorized modifications. The term is sometimes used in legal contexts to refer to the authenticity of digital evidence, where hash values are used to prove that evidence has not been tampered with.

Entity integrity

A rule that ensures each row in a table is uniquely identified by a primary key.

Referential integrity

A rule that ensures relationships between tables remain consistent, typically enforced through foreign keys.

Domain integrity

A rule that ensures data in a column falls within a defined set of valid values.

ACID

A set of properties (Atomicity, Consistency, Isolation, Durability) that guarantee reliable processing of database transactions.

Checksum

A computed value used to detect errors in data transmission or storage.