Data Integrity Checks
1. The Core Distinction
Data integrity is maintained through two distinct processes. It is vital not to confuse them in the exam.
Validation
Automatic computer checks to ensure data is sensible and follows specific rules before being processed.
Example: Ensuring a month is between 1 and 12.
Verification
Checks to ensure data has been entered correctly from the source or transmitted without error.
Example: Typing a new password twice.
2. Common Validation Checks
| Check Type | Description | Example |
|---|---|---|
| Range Check | Numbers must be within two boundaries. | Age between 0 and 120. |
| Format Check | Data must match a specific pattern. | Postcode (LLNN NLL). |
| Length Check | Minimum or maximum number of characters. | Password at least 8 chars. |
| Presence Check | Ensures a field is not left blank. | "Required" form fields. |
| Check Digit | A final digit calculated from others. | ISBN or Barcode numbers. |
3. Data Transmission Integrity
When data moves across a network, bits can be "flipped" (0 becomes 1) due to interference. We use these methods to detect errors:
1. Parity Check: An extra bit (parity bit) is added to make the total number of 1s either Even or Odd.
2. Checksum: A value calculated from the block of data before transmission. The receiver recalculates it; if they don't match, an error occurred.
3. Echo Check: The receiving computer sends the data back to the sender, who compares it to the original. (Inefficient but accurate).
2. Checksum: A value calculated from the block of data before transmission. The receiver recalculates it; if they don't match, an error occurred.
3. Echo Check: The receiving computer sends the data back to the sender, who compares it to the original. (Inefficient but accurate).
4. Methods of Verification
- Visual Check: The user manually compares the data on the screen with the original source document.
- Double Entry: Entering the same data twice into different fields. The system compares them and rejects if they differ.
⚠️ AS-Level Warning: Validation cannot prove data is correct. If a user enters "June" instead of "July", a format check will pass because "June" is a valid string. Only Verification (specifically a visual check) would catch this mistake.