Validation vs Verification: The Difference Students Keep Mixing Up
These two words cost marks every single exam series. They look similar, they both involve checking data, and they are almost always tested together — which is exactly why they get muddled.
So let us fix it with one clean mental model you can hold in your head:
Keep those two questions separate and the rest of this topic falls into place.
By the end of this lesson, you will be able to:
Validation: is the data reasonable?
Validation is an automated check that data is reasonable or acceptable according to a set of predefined rules, before it is processed. It does not — and cannot — check whether the data is true. It only checks whether it is allowed.
The main validation checks you should know:
| Check | What it does | Example |
|---|---|---|
| Range check | Rejects values outside an acceptable range | An age of 250 is rejected |
| Length check | Rejects data with the wrong number of characters | A password that is too short |
| Type check | Rejects data of the wrong type | Letters typed into an age field |
| Presence check | Ensures a required field is not left empty | A blank email field is rejected |
| Format check | Ensures data follows a required pattern | A date entered as DD/MM/YYYY |
Worked example. Someone enters an age of 250. A range check rejects it, because a human age of 250 falls outside any sensible range. The data was unreasonable, so validation caught it.
Verification: was the data entered correctly?
Verification checks that data has been entered or copied correctly — usually that it matches an original source. The two methods you should know:
- Double entry — the data is entered twice and the two versions are compared. If they differ, there is an error.
- Visual check (proofreading) — a person compares the entered data against the original by reading it carefully.
Worked example. The original email address is:
But it is typed as:
This might pass several validation rules — it has an @, it is a sensible length, it has the right format. Yet it is still wrong, because it was copied incorrectly. Only verification catches a mistake like this.
The idea that ties it together: valid does not mean correct
This is the sentence that unlocks the whole topic:
Validation only proves the data is allowed. It never proves the data is true. That gap is exactly why verification exists.
The strongest worked example
An examination mark must be between 0 and 100. A student actually scored 76. The operator mistypes it as 67.
67is within 0–100, so it passes the range check.- Therefore the value is valid…
- …but it is not correct. The real mark was 76.
Validation is happy. The record is wrong. Detecting the transcription error — 76 becoming 67 — needs verification, such as double entry against the original mark sheet.
Exam Tip: read the keyword in the question
The question usually tells you which one it wants.
If it says “prevent unreasonable data” → think validation.
If it says “ensure data has been entered correctly” → think verification.
Retrieval Check: validation or verification?
Decide which term applies to each scenario, then reveal the answers.
- A field rejects a phone number that contains letters.
- An operator types a customer’s address twice and the system compares both.
- A date must be entered as DD/MM/YYYY or it is rejected.
- A student proofreads a typed quotation against the printed original.
- A booking form refuses to submit while the name box is empty.
Check the answers
1. Validation (type check). 2. Verification (double entry). 3. Validation (format check). 4. Verification (visual check / proofreading). 5. Validation (presence check).
In real systems, you need both
Validation and verification are not competitors — they protect against different failures, so good systems use them together. Validation can run at the point of entry and block data that breaks predefined rules. Verification guards against the honest human slip: a correct-looking value typed in place of the real one. Remove either and a gap opens. That is why an exam answer that treats them as a pair, each doing a job the other cannot, tends to read as the strongest.
Final Summary
Validation
Automated check that data is reasonable / acceptable by predefined rules.
Verification
Check that data was entered or copied correctly (double entry, visual check).
Key Idea
Valid does not mean correct. Passing a rule is not the same as being true.
Exam Success
Match the keyword: “reasonable” → validation, “entered correctly” → verification.