By: Prof. Dr. Fazal Rehman | Last updated: November 12, 2022
Black Box – Robust Testing in software testing
In robustness testing, the software is tested by giving invalid values as inputs. Robustness testing is usually done to test exception handling.In robust boundary value testing, we make combinations in such a way that some of the invalid values are also tested as input.
The type of testing done by causing the software or system to fail in order to test the robustness is called robustness testing.
Robust Boundary value testing on 3 variables:
Suppose we have 3 variables X, Y and Z to testThe range of X:0 to 100The range of Y:20 to 60The range of Z: 80 to 100
X
y
z
Min-
-1
19
79
Min
0
20
80
Min+
1
21
81
Nominal
50
40
90
Max-
99
59
99
Max
100
60
100
Max+
101
61
101
Figure:Testing points detected in Simple Robust Boundary Value TestingTest Cases:Total Test cases = (Number of variables * Number of testing points without nominal )+ (1 for Nominal)These testing points are min-, min, min+, max- and max and max+19=(3*6)+1We can generate 19 test cases from both variables X, Y, and Z.
There are a total of 3 variables X, Y and Z
There are 6 possible values like min-, min, min+, max-, max and max+
1 is for nominal
Logic:When we make test cases, we will fix the nominal value of the two variables and change the values of the third variable.For example
We will fix the nominal values of X and Y and make a combination of these values with each value of the Z variable.
Fix nominal values of X and Y are 50,40, and we will compare these two values with 79, 80,81,90,99,100 and 101.
We will fix the nominal values of X and Z and will make a combination of these values with each value of the Y variable.
Fix nominal values of X and Z are 50, 90, and we will make a combination of these two values with 19, 20,21,40,59,60 and 61.
We will fix the nominal values of Y and Z and will make a combination of these values with each value of the X variable.
Fix nominal values of Y and Z are 40, 90, and we will make a combination of these two values with -1, 0,1,50,99,100 and 101.
Test Case#
X
Y
Z
Comment
1
50
40
79
Fix Nominal of X and Y
2
50
40
80
Fix Nominal of X and Y
3
50
40
81
Fix Nominal of X and Y
4
50
40
90
Fix Nominal of X and Y
5
50
40
99
Fix Nominal of X and Y
6
50
40
100
Fix Nominal of X and Y
7
50
40
101
Fix Nominal of X and Y
8
50
19
90
Fix Nominal of X and Z
9
50
20
90
Fix Nominal of X and Z
10
50
21
90
Fix Nominal of X and Z
11
50
59
90
Fix Nominal of X and Z
12
50
60
90
Fix Nominal of X and Z
13
50
61
90
Fix Nominal of X and Z
14
-1
40
90
Fix Nominal of Y and Z
15
0
40
90
Fix Nominal of Y and Z
16
1
40
90
Fix Nominal of Y and Z
17
99
40
90
Fix Nominal of Y and Z
18
100
40
90
Fix Nominal of Y and Z
19
101
40
90
Fix Nominal of Y and Z
Figure: Test cases generated in Robust simple Boundary Value Testing.
Video Lecture
Another Example of robust boundary value software testing
There are many benefits of robustness testing. Some of the benefits are mentioned below;
1. Better project analysis
Robustness testing means to increase the study of what has already been analyzed about your product. The robustness testing extends the area of testing of the previously tested software components. Robustness testing also test invalid values to satisfy the testing level.
3. Better design
The robustness testing result in more options and better software designs and it is completed before the finalization of the design of the product.
4. Achieve consistency
Robustness testing helps to increase the consistency, reliability, accuracy and efficiency of the software.
Frequently Asked Questions (FAQ)What does robustness mean in hypothesis testing?
Robustness is the strength of a tests, and procedures according to the specific conditions of the statistical analysis a study hopes to achieve the goals.
Suppose we have 3 variables X, Y and Z to test
The range of X: 0 to 100
The range of Y: 20 to 60
The range of Z: 80 to 100