What are you views with regards to writing you 'own' test requirements and Test Procedures. These then being QA'd by Senior tester. After these are QA'd then the tester who created the tests then execute their own testing?
The above has been the preferred method however, othere have the view that the person that writes the test scripts should 'not' execute the tests. This is a quality check in its own right.
Be interested in other poeple/testers views on the above?.
There are risks when tester executes their own scripts, because sometimes a tester will just get too close, and if one tester writes and one executes, you're getting more people involved, looking from different angles.
But, it's normally slower, because the second person has to read and understand the scripts before they can execute them.
Which risk is bigger?
Sometimes thoroughly deeply testing scope is more valuable, sometimes it's the breath of scope. So, it depends. Let me think of examples:
We were testing a grey-box decision rules engine embedded on a portal. Scripts and execution were done independently for the decision rules engine and with BVA covered 100% of the calculations and decision points in the engine (about 120 from memory) , because it had to be right and it had to be trusted. The rest of the portal, the data-entry, cross-browser testing etc the testers wrote and executed their own tests, because getting across as much of possible quickly was the best way to communicate it's current quality. We used those approaches because the required communication of the quality of the AUT was the grey-box was 100% tested and all defects fixed, and the portal had had a proportion of the scripts executed and a range of defects logged.
I have found that like most things, it depends. It depends on whether your tester is experienced and trustworthy or perhaps is new and still earning trust. It all comes down to risk and the cost of mitigating that risk. I have rarely been in a situation where the risk was great enough that I have not allowed a tester to execute their own tests. I have always implemented some level of peer review for the scripts and held testers accountable for their work so my level of confidence was normally high. That being said, my testers have always known there was a possibility that someone else may run their tests if we get in a situation where one tester is finished testing and we want to redirect that resource to assist others in executing their scripts.