Black box testing is a primary concept that is used in test design. As the tester has no knowledge about the subject under test (SUT), all the testing is happening with data: providing SUT with input data and analyzing the output data. Test data refers to the set of input values and expected outputs. This data should be maintained carefully to allow separate data management and data heuristics.
Test Design Techniques
The majority of test design techniques, namely Domain Partitioning (e.g., Equivalence Partitioning), Boundary Value Analysis, Decision Table Testing, and Combinatorial Testing, are data-driven.
Data Management Rules
The following rules will make your data management more effortless:
A test is a tool where data is the key. Properly exposed data allows people to understand what the test does.
Reusability
When data is separated, a single test case can be reused with different data sets, reducing the need to create multiple test scenarios for similar functionality.
Coverage
As more test data sets are added, the test coverage increases, leading to better validation of the application's behavior.
Maintenance
Data is the most changeable part of your test. Proper test data identification is an invaluable clue to updating them effectively to keep the tests alive. In many cases, when the application changes, only the test data needs to be updated rather than the entire test scenario.