5.3) Equivalence Partitioning and Boundary Value Analysis

Question 1

What is equivalence partitioning?

Answer 1

Equivalence partitioning is a technique for selecting specific test data when generating test cases at the function level. All possible input values that can be assigned to a function are divided into equivalence classes. The term equivalence class comes from algebra, where it approximately means a set of similar elements.

Question 2

How does equivalence partitioning work?

Answer 2

With respect to the generation of test cases, this means that all input values that result in a functionally identical system behavior are grouped into an equivalence class. When conducting the tests, not all possible input values must be tested: instead, only one arbitrarily selected representative from each respective equivalence class need be used. With this approach, the number of test data required per test case can be reduced quite significantly.

Question 3

Please list the steps when generating test cases.

Answer 3

• Step 1: First, the equivalence classes for each input parameter of the function or each GUI element in the dialog screen are determined and clearly identified. If necessary, the result of Step 1 can be documented in a table for clarity.
• ** Step 2**: Test cases that cover all valid equivalence classes are generated. The number of test cases required should be kept as low as possible. As a consequence, the combination of test data should be chosen in such a way that as many equivalence classes as possible are tested through as few test cases as possible.
• Step 3: Test cases are generated to cover all invalid equivalence classes. In contrast to the test cases for valid equivalence classes, the test cases for invalid equivalence classes only test one invalid equivalence class per test case. All other test data is taken from valid equivalence classes. Thus, the relevant invalid equivalence class can be more easily identified in the case of a failed test. If several invalid equivalence classes were tested in one test case, it cannot be said with certainty which of the invalid equivalence classes is the cause if a test fails.
• Step 4: Finally, specific input data is generated for each test case based on the selected equivalencies. A total of nine test cases were identified for testing the function for booking the conference room.

Question 4

What are the advantages and disadvantages of equivalence paritioning?

Answer 4

• Advantages: Equivalence partitioning is a simple technique for determining test data and test cases. It can be used for any technical or business function that anticipates input data. Its tabular representation enables test cases to be both generated with ease and documented intuitively.
• Disadvantages: However, business dependencies between individual equivalence classes are not taken into account; each equivalence class is essentially independent. Furthermore, equivalence partitioning is not suitable for components or systems whose behavior depends on defined internal states, for example, the states of business objects or system components.

Question 5

What is boundary value analysis?

Answer 5

Boundary value analysis, like equivalence partitioning, is a technique for selecting specific input data. The underlying assumption of boundary value analysis is the observation that software errors often lie in the boundary region that represents the transition from valid to invalid input values, i.e., at the boundaries of equivalence classes.