How to do Regression Testing in a SCRUM Project?

What is Regression Testing?

In simple words regression testing is testing again what has already been tested. This testing is done when a modification, code changes or a new functionality added to the system. We do this testing to ensure that the modification will not affect the existing functionality which was perfectly working fine earlier.

When to do Regression Testing?

Whenever a modification is done, it is not feasible to do regression testing. So we have to come up with a plan. If we follow agile we can break several tasks for regression within the Sprint and we can carry out regression testing based on the milestones defined within the sprint. (Eg: Milestones: A date when a specific functionality or two being released and tested). This will ensure the existing functions are working as expected despite there are modification done time to time. If Waterfall once a modification is done to an existing functionality , better to do regression testing for that affecting module only.

You will have to run through all the test cases that's been executed so far which is related to that module and do a high-level testing based on those test cases.

Some Regression Testing Examples (When we need to do regression testing?)

1. When new functionality is added to the system
2. When there is a Change Request
3. When there is a modification in code

However, it is up to the team to plan the regression testing properly without doing it all the time.

I have noticed some teams doing regression testing at the end of the UAT cycle saying they are covering all the end to end test cases. However, it is not how the way it should be done. The testing you do after an UAT cycle which caters end to end is basically system testing.

However different teams follow different methodologies and they have different definitions for these testing approaches. This way I mentioned is what I practiced for a long period.