Definition
A black box approach does not consider the internal structure, design, and product implementation. In other words, the tester does not know the internal functioning of the SUT (system under test).
The black box approach only evaluates the external behavior of the system. The inputs received by the system and the outputs or responses it produces are tested.
Key features:
- considers only external behavior
- carried out by testers (with an end-user perspective)
- no need to have knowledge on specific programming languages or implementation details
- it exposes ambiguities or inconsistencies in the requirements specifications
- used in system testing or acceptance testing
- less time consuming compared to white-box testing
- it focuses on the behavior of the system
- it is not suitable for algorithms testing
Exhaustive input testing (p.9, (Myers et al., 2011)): making use of every possible input condition as test case. Since you cannot test a program to guarantee that is error free, the ultimate goal is to find as many issues as possible given time and resources constraints.
Black-box techniques
- EPM (Equivalence Partitioning Method)
- boundary value analysis
- cause-effect graphing
- error guessing
- decision table
Also known as: data-driven testing, functional testing, closed-box testing, input-output-driven testing