Test Republic

Community of Software Testing Professionals

It is used to test interactions between boundaries of fields with ranges, Normally the Boundary value testing is used for single data validation with seven values like Min-1,Min,Min+1,Mid,Max-1,Max,Max+1, In this Boundary TestCase Generator we can generate multiple interactions of different fields range.

For Example:
Range1: 10-45
Range2: (-40)-(-50)
Range3:1000-5000
Testing manually single data is good but testing the interactions is more reliable for that we have to go to the tool which will reduce our work as simple as possible.
Interacting Min Value of Range1=10 with Range3=5000, If it passes when the testing has been done for single Range value, there is a chance of occurring the bug when the same value interact with another Range value.
Total possible Test Combinations=7*7*7=343
If we go through the tool it reduces into 53 combinations which will cover almost all the combinations.

Reply to This

Replies to This Discussion

You can also try using PerlClip for boundary value analysis. KIndly ignore this reply if you already know about PerlClip.

Reply to This

Santhosh,

Perlclip is not a Boundary Value Analysis tool. The BVA++ of Testersdesk has a unique purpose in addressing the Boundary Value Tests, I want you to figure it out and let me know in the reply why I might be right or wrong.

Reply to This

Hi Santhosh,

What pradeep has told is exactly wright; first of all please go through the Boundary TestCase Generator in TestersDesk.com you will get the clear information about the normal BVA++ and to this BVA++

Reply to This

Suresh, Can you publish a case study of BVA++ here to help Santhosh and others understand it?

Reply to This

HI
Case Study for Boundary TestCase Generator

Here I am going to explain simple case study for BVA++, to easy understand we can take our Case Study itself, about the employees
Age: 24-54
Salary: 8000-20000
Employee Type :{ Tester, Developer, Manager}
Normally the Boundary value testing is used for single data validation with seven values like Min-1,Min,Min+1,Mid,Max-1,Max,Max+1, In this Boundary TestCase Generator we can generate multiple interactions of different fields range.
EG:
For Age the boundary values are 23,24,25,39,53,54,55
For Salary the boundary values are 7999, 8000,8001,14000,19999,20000,20001
For Employee the possibilities are Tester, Developer, Manager
TC1:
From the above example minValue of Age 24 can interact with maxValue of Salary 20000
If For single Data validation the minvalue of Age 24 will pass and there is a chance of occurring the bug when this minvalue is interacts with another maxvalue or Max+1 value
TC2:
Max Value of Age 54+ Mid Value of Salary 14000+Manager
Like above TestCases we can write multiple interactions.

Reply to This

Take a look once again at the below Test Input Model-

Age: 24-54
Salary: 8000-20000
Employee Type :{ Tester, Developer, Manager}

From the test coverage perspective - we have 7 values for "Age" to be tested (min-1,min...) using normal BVA, 7 values for "Salary", and 3 values for "Employee Type". So would 7 + 7 + 3 (= 17) test configurations be sufficient?

NO. 7 * 7 * 3 (= 147) tests is what it takes for exhaustive coverage of all types of interactions between all the fields.

Our Boundary TestCase Generator (BVA++) can reduce the 147 combinations/configurations to just 49, ensuring that all pairs of values are covered.

Traditional BVA talks of only one field at a time without considering the other test parameters that could influence the outcome, and our BVA++ extends that scenario by providing Pairwise Coverage on the group of range based (including list based, as you can see in "Employee Type") input fields.

Reply to This

Traditional BVA talks only of the boundaries of which you've been told. The kind of boundary testing that we teach (and that we encourage all testers to think about) talks about discovery of boundaries that other people haven't considered.

---Michael B.

Reply to This

Very true.

We teach the postulate that a boundary can be any point disposition in program logic, known, or unknown. And ways how to analyze such things in a context-specific manner.

The tool we provide applies to the treatment of boundaries as the numeric edges and the values on both the sides of the edge (which is what people have been told for decades). However, we do go a step ahead to create pairwise interfactions between such boundaries, thereby achieving more coverage compared to the single factor coverage.

The recipe we provide is the chart we produce (automatically feeded into a tool based test script that runs the tests), and use the "human time" to think and exercise the other "possible" boundaries that operate in the specific context.

Reply to This

RSS

Test Republic Elsewhere

 /></a></p> <p style=

Members

  • jagan
  • P.K.Ramya
  • Sivakumar
  • Arun Kumar.V
  • Poornima Kadambi
  • Vijayalaxmi
  • Sreethin
  • Brian Osman
  • s kumar
  • Shreya
  • APARNA
  • Sanjeev Kumar Singh
  • BIDISHA BAGCHI
  • Ipsita ratha
  • Sapna Nair

© 2010   Created by EDISTA.

Badges  |  Report an Issue  |  Terms of Service