Recently I’ve been thinking about how better to talk about test types and what they are trying to convey, as I’ve never really felt test types and levels and methods are all that useful.
So what I did was go back to basics. What is testing all about? Quality. Risk. Information. Value.
Then it struck me, testing types are all about risk. If our job is to provide information about quality and risk, then the testing types we tend to use to frame discussions should relate to quality and risk too.
I started thinking about a model that links risks (either known or to be identified) and although there are many typical risks that exist that easily map to a testing type, for example the risk of something not working across different browsers and the compatibility testing type, there are many too that don’t easily map to a testing type.
Or more importantly there are many risks out there worth exploring that aren’t looked at because we tend to pick a few types of testing that limit the scope, breadth and depth of the testing we do.
It was about this time that I stumbled across a superb blog by Dan Ashby that explained a bit more about the idea with a bit more substance and examples.
I’ll leave you with that as any more feels like I’m stealing someone else thunder. Enjoy.