Here is the comments and discussions on my previous question in a famous qa forum

Jerry-
At last a Software Test Analyst's effectiveness is just based on his number of Defects….? This may looks funny, but is the Truth. In a project a Software tester's effectiveness is just measured (at last...) by his number of defects.This happened in my nearby project. The team has to release a resource according to the plan and the team manager is breaking his head to make a decision. At last he got two names to select one. Member A is very good in overall application knowledge and designing test cases. Member B is very good in finding defects (even apart from TC execution), with lot of sense and logical aptitude. Now almost the TC design stage is over and they are in the execution stage. What the Team Manager do, whom will he release...Pls post your comments.

Muthu-
Hi Rajan,I would recommend retaining Member B since your going to start with Execution Cycle. Test Case updating would be minimal during this stage.As a Test Manager we would need a Resource with good Logical and Analytical Skills to find more Valid Defects.

Richard-
Member B does sound like the logical choice, however the Test Manager must ensure that they do not lose any domain knowledge that can not be gained elsewhere. So they need to check that any information member A has can also be gained from other team members or get them to document it.

Jerry-
Thanks Muthu and Richard,Both views sounds good.If he released Member A: 1. The team manager will lose a domain knowledge Resource.2. No one is equivalent and accurate like executing his own test cases. (Hope we all experienced when executing other's cases)3. In India we use a term "Karuveapili" (the ingredient we add in food for good flavour, once cooked we throw it away)the Member A must not treated in such a way. The manager has to give him some acknowledge to that member and retain him in the project.4.I personally know that even Member B find some good critical logical defect apart from Use Cases he used to consult and clarify that defect with Member A.(I'm very confused)

Richard-
Sounds like there is not going to be an answer that is perfect, the project is going to lose out some way, and somebody is going to be upset. (assuming that they are not leaving the company I would do the below)I would explain to member A that he has to lose one person. Make it clear that they have done tremendous work on the project, and the project would not be in such a good state going into the testing phase without them, however he is letting them go as they are much more valued for their test case planning and use case reviewing (etc etc) and therefore will be more value to another project in that capacity. To member B I would explain that the have been kept on due to their excellent eye for detail and finding really good incidents. However as member A is moving on to other projects this is a really good opportunity for them to grow in their role and take responsibility and ownership of the issues and have the confidence to speak to developers/designers etc to clarify issues. Hopefully this will mean member A understands that they are not Karuveapili (love that phase, thanks!) and member B can grow in their role and it will effect the project in a good way. Does that help?

Jerry-
Wow Man Richard,Really wonderful decision. I'm feeling very happy on reading this reply. (Small fish with big humanity and human administration...!)Many Thanks for your effort.

Joe-
I completely disagree.Defect count by itself is a poor measure of effectiveness.Consider...Tester 1 finds 100 defects. All of them are surface-level defects in code that has many, many defects present. And, the existing critical defects have been completely missed.Tester 2 finds 10 defects. All of them are critical, and there are no other critical-level defects remaining in the code.I know which of the two I would choose.

Jerry-
Yes Joe, you r right.In your situation, Every one will give First priority to the Tester 2 and Second priority to Tester 1.(Anyway a defect is a defect is a defect, doesn't matter big or small)But here the case is different. Either you keep the one good in finding defects or the one good in Domain Knowledge and talented in Test Design.


Hope you all enjoyed.
-Jerry