7.6 Test Data

(a) Describe what is meant by test data:

Test data is data that has been specifically created and selected for testing a program. It is used to check whether a computer system works correctly and to identify any errors (bugs) in the code before the program is released for real use.

(b) Difference between normal, abnormal and boundary test data:

(c) Three reasons why test data is used:

• To ensure the program works as expected for valid input.
• To find errors (bugs) in the code before the program is released.
• To check robustness — to ensure the program handles invalid input safely without crashing.

(a) Example of normal data:

Any whole number between 11 and 19, e.g. 15. (It is strictly inside the valid range — not on the boundary.)

(b) Example of abnormal data:

Any value outside the valid range or of the wrong type, e.g. 9, 21, 15.5 (non-integer), or "ten" (wrong data type). All these should be rejected.

(c) Example of boundary data:

The two values that sit exactly on the edges of the valid range — 10 and 20. Both should be accepted. (The values just outside — 9 and 21 — are also useful boundary tests that should be rejected.)

(a) Two items of normal data:

Any two values strictly inside the valid range, e.g. 20 and 30. (Both should be accepted.)

(b) Two items of abnormal data:

Any values clearly outside the valid range or of the wrong type, e.g. 10 (below 16), 70 (above 65), -5, or "ten" (wrong data type). (All should be rejected.)

(c) Four items of boundary data:

The two values on the edges of the valid range and the two values immediately outside them — 15 (rejected), 16 (accepted), 65 (accepted), 66 (rejected). Testing both sides of the boundary confirms whether < vs <= has been written correctly.

(a) Test data table:

(b) Why both sides of the boundary must be tested:

Testing both 20 (accepted) and 21 (rejected) confirms that the programmer has used <= rather than < in the selection statement. If only one side is tested, an off-by-one error at the limit may go unnoticed — the program might wrongly accept 21 tickets or wrongly reject 20.

(a) Range of valid numbers accepted:

Any number greater than 0 and less than or equal to 100 — i.e. 0 < Number ≤ 100. The lower bound 0 is excluded (because the test is > 0), but the upper bound 100 is included (because the test is <= 100).

(b) Example of abnormal data:

Any non-numeric value such as "hello" or "abc". The program would crash or produce an unexpected result because the input cannot be compared with a number.

(c) Boundary data that produces "Valid":

100 — the upper boundary, which satisfies Number <= 100. (The value 0.01, just above the lower limit, is also a valid boundary test.)

(d) Boundary data that produces "Invalid":

0 — the lower boundary, which fails the test Number > 0. (The value 101, just above the upper limit, is also an invalid boundary test.)

(a) Boundary test data:

Any value that sits on the length limit or just outside it, e.g.:

• Pass1234 — exactly 8 characters (boundary, accepted).
• Pass123 — 7 characters, one short (boundary, rejected).
• Pass12345 — 9 characters, one long (boundary, rejected).

(b) Test data table:

(a) Two sets of boundary test data:

• Set 1: Age 15 ($5) and age 16 ($10) — the boundary between the Child and Adult ticket prices.
• Set 2: Age 59 ($10) and age 60 ($7) — the boundary between the Adult and Senior ticket prices.

Each pair tests both sides of the boundary — one value that should fall into one price band, and the value immediately next to it that should fall into the next price band.

(b) Why boundary data is important:

The ticket price changes at the boundaries 16 and 60. If the programmer wrote age > 16 instead of age >= 16, a 16-year-old would be charged the child price ($5) instead of the adult price ($10). Only boundary testing will catch this kind of off-by-one error — normal data such as age 25 would still produce the correct answer even if the comparison operator were wrong.