Raju Mandapaty

• Nov 22nd, 2006

 

We have to take 20% of total number of testcases for regression testing. This 20% of testcases should be selected randomly.

srinivasulub1981

 Profile Answers by srinivasulub1981 Questions by srinivasulub1981

• Nov 27th, 2006

 

HI

   Actually regression testing is done in two situations.. one is after bug fixing and another is whenever a new CR is came and have to test existing functionality once again if it reflects with new modifications..

   So depends upon these situations and priority, ( severity if needed) we can select the specific testcases from test build..

Srinivasulu.B( srinivasulub1981@gmail.com)

walkinwaz

 Profile Answers by walkinwaz 

• Nov 28th, 2006

 

Determine the type of testing, new feature, business rule change, or a defect correction.

I use the documentation provided by the developer that tells me what has been changed, added or corrected and how they are accomplishing this.  If a business requirement document is available, I use that to make sure the developer is meeting the requirements.

Then I create a query against the test data (if you have pre-existing data and do not need to create it) to find test cases that will test the change, additions or corrections.  The query will also help me select cases that make sure the existing functionality is working as expected.

naresh reddy

 Profile Answers by naresh reddy 

• Jan 19th, 2007

 

hi ,
this question can be answered in 2 ways.
1.we select the  tesr cases that effect the functionality of the application and then we regression test on those.
2.we give the priority while writing the test cases and in the regtression test phase we test the cases taht has the highest  priority.
if u r in product  testing  every test case is important. and every thing is  retested.
Regards,
Naresh Reddy,

rajeevn

 Profile Answers by rajeevn 

• Jun 16th, 2008

 

Regression test are prioritized as per the last code modifications.

As per the last code modification, you need to make a list of dependent test cases that may be required to test so as to check that the dependent componets are working as they were in the past.

So choice can be made only after carefully evaluating the code change and dependency analysis.

kurtz182

 Profile Answers by kurtz182 Questions by kurtz182

• Dec 11th, 2009

 

From a test repository of 1500 test cases, I would select the most vital 100 using the following guidelines:

1) What functional areas have exhibited the greatest vulnerability throughout the software development life cycle?  I would select test cases that can potentially identify these risks.

2) What functional areas have just been modified by bug fixes?  I would select test cases in and around these areas.

3) What functional areas are impacted by change requests?  I would select test cases in and around these areas.

4) What functionalities are dependent on high risk areas, areas modified by bug fixes, or areas impacted by change requests?  I would select test cases that verify these functionalities remain intact.

5) I would include a spattering of high level test cases in core mission-critical areas. 

6) In all cases, I would select high-level functionalities rather than low-level functionalities in each of the target areas.