When QA or UAT comes, should the developer still be involved in setting up data or finding test data for the QA tester or business user? Or will this introduce bias to the developer who coded the system changes that are being tested in the said environments?
- Anybody can ask a question
- Anybody can answer
- The best answers are voted up and rise to the top
While developer bias with creation of the test code is a risk, many times the developer knows how to do it more quickly than the QA resources could. I have set up test data and environments for QA resources before and it tends to work out in one of two ways:
In other words, if they NEED you to do this then it is a bad idea, if they WANT you to do this, then it is probably okay.
Depends heavily on what kind of resources/people the organization has access to, and it depends on the people themselves.
Some organizations can't afford not to utilize developers in this area. And some developers are as good or better at QA then the QA people are, while others are lacking in this area. There's no reason to make a blanket statement and say it should be one way or the other.
You want to have a good QA person/team. If your developer has this skill then that's a bonus you shouldn't pass up.
You should weigh the bias against the potentially saved labor. The developer brings a good amount of knowledge about the system and potential test cases, and can certainly help and shorten the process. I would normally put the burden of avoiding bias on the QA or business person - it's part of their job.
However, this depends for a good part on the actually involved persons. With the wrong combination of personalities, group-think becomes a major issue and the developer will taint the testing with too much of a developer's perspective.