Test data and test plans

Build the idea

Start with the concept and its key vocabulary.

Key idea

Select specific test values, predict their expected results and record them in a test plan.

Good testing is destructive: it tries typical values, edges and misuse to find hidden faults. For an inclusive range 1–5, 3 is normal, 1 and 5 are valid boundaries, and 0 or 6 are invalid. Text such as bananas is erroneous when an integer is expected. A test plan states the exact input and expected response before recording what actually happened. Iterative testing checks modules during development; final or terminal testing checks the complete program at the end.

Normal test data

Typical valid input that should be accepted and processed correctly.

Boundary test data

Input at the very edge of an accepted range.

Invalid test data

Input of the correct type that lies outside the permitted rules or range.

Erroneous test data

Input of the wrong data type.

Test plan

A record of test data, test type, expected result and actual result.

Iterative testing

Testing modules during development so faults can be corrected as the program is built.

Final testing

Testing the complete program at the end of development before release.

Keep these in mind

Normal data is typical and should be accepted.
Valid boundary data uses the exact accepted endpoints.
Invalid data has the expected type but breaks a rule.
Erroneous data has the wrong type entirely.
Test plans include input, category, expected result and actual result.
Testing aims to find errors, not merely prove that one example works.
Iterative testing happens during development; final testing happens after the complete program is built.

Select specific test values, predict their expected results and record them in a test plan.