Boundary values (Boundary Value Analysis, BVA) and equivalence partitioning (Equivalence Partitioning, EP) are fundamental test design methods in manual testing.
Background:
The techniques emerged to reduce redundancy in test scenarios and increase the likelihood of finding defects with less effort. Equivalence partitioning allows for dividing all possible input data into groups that should be processed the same way, while boundary values identify bugs that often occur at the edges of ranges.
Problem:
The main difficulty in applying these techniques is the incorrect definition of boundaries and partitions. For example, testers may misdefine ranges (e.g., by one unit) or overlook implicit boundaries (e.g., will 0 work if the range starts from 1).
Solution:
Effective application requires carefully reading the specification, clearly defining what input data is expected, and accurately marking closed and open boundaries. It’s always advisable to discuss ambiguous cases with analysts or developers.
Key features:
If there are many equivalence classes, should all their boundaries be tested?
No, it is necessary to highlight the most critical ones for the business (only valid and invalid) and not overload the test matrix.
Does the boundary value itself fall within the range?
It depends on the conditions of the task: if the range is [1, 10], then 1 and 10 are included; if (1, 10), then they are not.
Can these techniques be applied to text and string data?
Yes, for example, to check string lengths, permissible characters, and null values.
A tester, when checking the "Age" field (1-120), only created test cases for values 20, 50, and 100.
Pros:
Cons:
The tester checked all boundaries: 0, 1, 2, 119, 120, 121, and a random value within the range.
Pros:
Cons: