Data-Based Test Techniques

In the ISTQB Certified Tester Advanced Level Test Analyst syllabus (v4.0), data-based test techniques are black-box techniques that derive test conditions and test cases from the data a test object processes: its inputs, outputs, configuration parameters and internal values. They help the Test Analyst choose a small, representative set of values from a usually huge or infinite data space, while still giving good coverage and a high chance of finding defects. The syllabus covers three main data-based techniques. First, domain testing generalizes equivalence partitioning and boundary value analysis. The input (and output) domain is split into subdomains whose values are expected to be processed in the same way. Boundaries between subdomains are then tested carefully, because defects often occur where conditions are wrongly implemented, for example using '<' instead of '<='. Domain testing also covers domains defined by several variables at once, where boundaries may be expressed as relationships between variables, and it uses ON, OFF, IN and OUT points to target likely boundary defects. Second, combinatorial testing handles test objects with many parameters, each with several possible values, where testing all combinations is impractical. Techniques such as pairwise testing, each-choice testing and the classification tree method systematically reduce the number of tests while still covering every value pair or every chosen value at least once. This is based on the observation that most failures are triggered by interactions of only a few parameters. Constraints between parameters must be considered so that impossible combinations are excluded. Tools are commonly used to generate the combinations. Third, random testing selects input values randomly according to a chosen distribution, often an operational profile. It is useful when large numbers of tests can be run automatically and a reliable test oracle exists, and it can find unexpected defects that systematic techniques miss. The Test Analyst must select appropriate techniques based on risk, the nature of the test object and the required coverage, and must be able to measure and report the coverage achieved. In the ISTQB Certified Tester Advanced Level Test Analyst syllabus (v4.0), data-based test techniques are black-box techniques that derive test conditions and test cases from the data a test object processes: its inputs, outputs, configuration parameters and internal values. They help the Test Anal…