A Guidebook to Regression Testing

Regression testing involves re-running functional and non-functional tests. Regression testing ensures that the previously tested software performs well post any code changes in the software. It checks if the previous functionality of an application continues to work even with the addition of a new feature. It also checks for any new bug or error in the existing software.

Whenever a new software is released, there is a need to test it for its new functionality. The old tests need to be re-run to make sure that the new software does not re-introduce old defects nor does it create fresh defects. The regression testing is thus a verification process that gains maximum significance when there are continuous changes and improvements made in the software or application.

When should one perform regression testing?

1. New functionality: When a new code has been introduced by a developer, its compatibility is not tested. The software should be tested for any likely bugs or issues after the addition of new code.
2. Modification in the existing functionality: When developers revise existing functionality and edit or exclude or include features as per requirements, regression and functionality testing is to be done to check if the changes have in any way impacted the existing functionality or not.
3. New integrations with other products: When there are new integrations to be made with other products, regression testing helps ensure that the integrations are smooth.

Regression testing can be done manually or through an automated regression testing framework. If the software undergoes frequent changes, then automated regression testing would be the better option than manual testing. 

The various methods of regression testing are:

1. Retest all
2. Regression test selection
3. Prioritization of test cases

Retest all: In this method, all the test cases in the suite are to be re-executed. While it ensures that there are no errors, it is expensive as it consumes a lot of time and resources.

Regression test selection: In this method, only select cases from the test suite are run to check the impact of code changes on the software. This is because the cost of retesting all methods is very high.

Prioritization of test cases: In this method, test cases are scheduled such that those in higher priority get tested first and those in lower priority get tested later. There is general prioritization and version-specific prioritization. Prioritization is based on business impact, critical functionality and frequently used functionality.

Choosing the test cases for regression testing can be quite challenging. It is, indeed, an art. Some of the test cases that gain priority over others are:

• Test cases with frequent defects
• Functionalities that are more visible
• Test cases of functionalities that have undergone more changes
• All Integration Test Cases
• Test cases that verify the core features of the product
• All Complex Test Cases
• Test cases of functionalities that have undergone recent changes.
• Successful test cases – a sample 
• Failed test cases – a sample

Important tools for regression testing are – 

1. Avo Assure
2. Eggplant
3. Selenium
4. Quick test professional

Major testing challenges while performing regression testing

• When there are successive regression runs, test suites become large, making it expensive in terms of time and resources for the entire regression suits to be executed.
• It is challenging to minimize the test suite while gaining maximum test coverage.
• Determining the frequency of Regression Tests, whether after every single modification, build update, or after bug fixes is a challenge.

An effective regression strategy saves both time and money for organizations. Regression testing is used not only to test the correctness of a program but also to track the quality of its output.