Multiple choiceQuestion 1 · Recall
Loading…
By the end you can explain what validation and verification each check, why a system needs both, and why neither one guarantees that stored data is true.
A database only looks authoritative because it is printed neatly.
If a clerk types 19/07/200 instead of 19/07/2001, the record is wrong forever — and every report built on it is wrong too. Computer people summarise this as garbage in, garbage out.
Integrity (the accuracy, completeness and consistency of data) is protected by two guards, and they answer different questions:
The order matters. Validation runs first, as the data is entered.
Verification runs alongside or afterwards, to confirm nothing was mistyped or corrupted.
Data validation
Data validation is an automated check, performed by the system, that data entered is sensible and acceptable for the field before it is stored. Data that fails validation is rejected and the user is asked to correct it.
It does not check whether the data is true. A date of birth of 01/01/1900 may pass every validation rule and still be wrong.
Data verification
Data verification is checking that data has been copied or transferred accurately — that the value now in the system is the same value that was in the source.
Verification is about faithfulness to the source, not about whether the source was right in the first place.
Case 1. A patient's date of birth is typed as 31/02/2000. February never has 31 days, so a validation rule rejects it immediately.
The data never reaches the database.
Case 2. A customer's date of birth is 03/04/1999 (3 April). The operator types 04/03/1999 (4 March).
Both dates are real, so validation accepts it — there is nothing unacceptable about 4 March. Only verification can catch this: a second operator types it again, or someone compares the screen with the paper form.
That is why both exist. Validation catches impossible data automatically and cheaply. Verification catches plausible-but-wrong data, which needs human or duplicated effort.
Both guards protect the copying of data. Neither checks the real world.
This is worth a sentence in an exam answer: validation and verification reduce errors, they do not eliminate them.