I have recently moved to a unit where my role is to provide test data to testers and create reports in oracle.
The company is rolling out a new n-tier application and it needs to be tested.
The lead of my group is a Business Analyst with no knowledge of what it will take deliver on our tasks.
He wants me to minimize the analysis on the new data model and focus more on tasks like creating scripts to get the data to the testers.. huh?
I have raised a couple of issues with him already but still he does not see the light.
1 - I mentioned that without complete and thorough analysis of the data model, we cannot give the testers test data that they require. We also cannot create reports without knowing the data model inside and out.
2 - I also mentioned that since the data model that we have is not the finalized data model, creating scripts is a waste of time...
Opinions.. comments... suggestions on a site where I can get a detailed step by step data model analysis schedule or something.. At this point I just need to make sure that I get ready for a barrage of questions on why I cant give him the data he needs.. .
I though of saying that I broke my magic wand.. but I though that comment might get lost in translation..
Your manager is looking for "black box" testing, and you're trying to do "white box" testing. These are two fundamentally different kinds of tests, and are intended to produce two related, but quite different results.
A "white box" test like you are trying to do involves knowing exactly how the system is built, and testing all of the possible constructions of data. This exercises/tests the data model, and is the second highest commonly used form of testing. This takes a lot of effort, full knowledge of how the data is stored, and special cases that need to be checked.
A "black box" test is done at a much higher level. It tests the UI, looking for common problems, often concentrating on the details of the UI without worrying about the details of the system that underlies that UI. This is a very high level test and is no where near as thorough, but it takes much less effort.
The short answer is that you are trying to do a much lower level test than what your manager feels is necessary or cost effective. It is always a great thing to strive for more than is expected, but it probably isn't going to be an easy sell, since your manager already envisions a much higher level test than you are planning!
Since you work for this manager, he (not you) gets to call the shots. You are a professional employee with significant expertise that he doesn't have, and as a good employee you should make use of your expertise. I wouldn't "dig my heels in" for one approach or the other. Rather, I would advise the manager of the pros/cons of each approach as appropriate, perhaps even in memos or e-mails (so you have documentation). Then, do what he asks you to do to the best of your ability.