Sunday, October 7, 2018
The six presentations at CEWT #6 took different angles on the topic of good testing or good testers. Here's my brief summary of each of them:
Whatever good testing is, Karo Stoltzenburg said, it's not likely to be improved by putting a box around its practitioners. In fact, pretty much any box you care to construct will fit some other profession just about as well at it fits testing. Testers, like people in general, are individuals and it's in the combination of diverse individuals that we spread our risk of missing something and so increase our chances of doing the good testing we all want.
What makes a senior tester? That's the question Neil Younger has been asking himself, and his colleagues at DisplayLink. He shared some of his preliminary thoughts with us. Like Karo he wants to avoid boxes, or at least to reduce their rigidity, but against that there's a desire from some for objective criteria, some kind of definition of done that moves them from one box to another. A senior tester for Neil must naturally also be a good tester, and the draft criteria he shared spoke to the kinds of extras he'd expect such a tester to be able to provide. Things like mentoring, coaching, unblocking, improvement, awareness, and knowledge.
"Any kind of metric for testing that involves bug counts focusses on how we're sorting out the failures we find, not on team successes." Chris Kelly has seen metrics such as net promoter score applied to technical support and was musing out loud about whether metrics could be applied to testing. If they could, what should they look like? The discussion covered the difference between teams and individuals. Can a team member do good testing while the team fails? One way of judging success was suggested: ask the stakeholders whether they're getting what they want from your work.
In the limit, good testing would lead to perfect testing, Šime Simic proposes, perhaps a little teasingly. We aren't going to get there, naturally, but we can do things that help us along the road. In particular, he talked about knowing the mission, aligning the mission with stakeholder needs, testing in line with the mission, doing what we can inside the constraints that the project has, and then reflecting on how it went and how we might change next time in similar circumstances. Testing, he went on, should not exist in a vacuum, and it's unrealistic if not unfair to try to judge it apart from the whole development process. (References)
Because they have different needs it's perfectly possible, and indeed reasonable, for two people to have different views of the same piece of testing. When they are the test manager and the tester, however, it may lead to some problems. For the tester, on the project Helen Stokes and Claire Banks talked about, exercising the product was the primary requirement. For the manager, visibility of the work, and the results of the work, were imperative. "There's more to good testing than doing the tests" they conclude.
My own presentation was about how, assuming we could know what good testing is, it can be a challenge to know whether someone is capable of doing it. I talked about how this particular problem manifests in recruitment all the time, and about how testing skills can be be used to assess candidates for testing roles. (Slides)