Take the 2-minute tour ×
Software Quality Assurance & Testing Stack Exchange is a question and answer site for software quality control experts, automation engineers, and software testers. It's 100% free, no registration required.

Currently most of the industries are using Agile based software development methodologies. And in Agile based development, we already have test cases in hand before development work starts. So, can anyone tell me how software defect prediction model fit/help in this type of development from industry's point of view ? (Mean reduction in testing cost or improving software quality).

There is one more thing. In agile based development, QA activity is performed by developer. So, can the developer use the software defect prediction at his end for checking the code for bugs? or in a agile project, at the end of iteration when a functioning piece of working software is released, we can perform defect prediction to find out defect-prone files.

I am trying to find out how traditional software defect prediction process can be use in agile environment?

share|improve this question

2 Answers 2

I don't think it's much use - the idea of agile is rapid feedback and so bugs are found quickly not 6 months later using defect prediction. I'm also dubious about using default prediction models anyway

share|improve this answer
Can you explain more about "I'm also dubious about using default prediction models " –  Rajendra Jun 12 '14 at 5:13
Because "it depends". Capers Jones can probably trot out studies that show 1000 lines of code will have 2.5 defects - but it depends on the project, the experience of the devs and so many other factors that I find it useless. –  Phil Kirkham Jun 12 '14 at 13:55
But isnt it the case that defect prediction process takes into account project complexity, the experience of the devlopers as well as testers, resource constraints etc? 2.5 defects per 1000 line might be industry average but every project has its own characteristics. Take the analogy of building bridge. One can argue that studies/stats shows to build a bridge, it needs “X“ amount cement per 100 meters. But while building new bridge one has to take in to account factors like height, flow /speed of water etc and predict the amount of cement. –  Rajendra Jun 23 '14 at 10:03
Please, no bridge building analogies :( –  Phil Kirkham Jun 23 '14 at 11:48

One can think defect prediction is a tool in hands of project owner (who has ordered to execute this project and given funding etc) which is indicator of

if there is some major goof up in development process or

still how many bugs should be expected from customer or

how effective current testing/test planning is.

In this sense predication is still relevant. However on its own defect prediction cannot do reduction in testing cost or play a role in improving software quality.

Of course one has to keep in mind that software developments is still a non-exact science and hence some amount of heuristic plays a role in predication.

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.