Software Test Services: Boundary Values Method

The given black box method uses Boundary Values. Input or output boundary value is on the edges of the selected equivalent area and, as a rule, are the minimum and maximum values of the area. 

When performing Boundary Value Analysis, test or test scenarios are created based on the chosen boundary values for input or output data. The given method complements the equivalence partition method.

Boundary Value Method is Characterized by the Following Definitions:

- Boundary values for "acceptable" equivalent area referred to as "permissible threshold values".

- Boundary values for "unacceptable" equivalent area referred to as "Invalid threshold values".

Tests are developed to cover both the boundary values of the permissible and not permissible values.

When Designing Tests by Means of This Method, Testers Use:

- 3 Value Boundary Approach. Boundary values are three numbers: boundary value itself and coming to him left and right, taking into account the multiplicity of values:

• with multiplicity equal to 1, for edge number 7, testers choose for the tests the number 6, 7 and 8;
• with multiplicity equal to 0.1, for edge number 7, testers choose he numbers 6.9, 7 & 7.1.

- 2 value Boundary Approach. For the boundary value, testers pick up 2 numbers, taking into account the multiplicity; there are left hanging, or right hanging approach to border:

• with the left approach to the border and when multiplicity equal to 1, for edge number 7, tester chooses perimeter number 6 & 7 for tests;
• with the right approach to the border and when multiplicity is equal to 1, for the edge number 7, testers choose 7& 8 numbers.

In practice, it is better to use an approach with three values at the border, because immediately achieved maximum coverage and defect are detected more often. 

This method can be applied at all levels of testing, taking into account input and output criteria cover.