Black-box testing involves viewing a system as if it is an opaque (black) box where only the inputs and outputs can be seen. This is a technique that is typically applied to overall system (high level) testing.

Suitable test data needs to be developed in order to carry out black-box testing. The group of all possible input values that can be inputted into a system is known as the input domain. All the possible inputs can be broken into categories (or sub-domains) that cover the testing requirements. The test data used will have at least one value from each input sub-domain.

Partition Analysis

Partition analysis aims to partition the input domain into sub-domains so that all the values within a single sub-domain will result in the same behaviour from the system. This means that only one value per sub-domain is required in testing because that one value should cover the effects of all the other values in the same sub-domain.

Partition analysis can be used to check that all of the functionality specified in the system requirements has been implemented correctly. Black-box testing has an advantage in this respect over white-box testing.

Boundary Value Analysis

Boundary value analysis aims to produce tests near the boundaries of sub-domains rather than just anywhere within the body of the sub-domains. Tests near the boundaries of sub-domains can be useful as failures often occur near the boundaries of input sub-domains.

Search for "Black-Box Testing, Partition and Boundary Value Analysis" on: Google | Kelkoo | Amazon | eBay (UK) | eBay (US)

Search for "Black-Box Testing, Partition and Boundary Value Analysis" on the rest of Computing Students: Black-Box Testing, Partition and Boundary Value Analysis