Sunday, 5 October 2014

Let's Test Oz

My third and final Let's Test Oz post; three experiences that left a lasting impression.

Cognitive Dissonance

Margaret Dineen gave a presentation where she spoke about cognitive dissonance*, the mental stress that people experience when they encounter things that contradict their beliefs. 

As an example, you might believe that your test environments have been configured correctly and continue to believe this despite repeated evidence to the contrary, perhaps because your test environment administrator is usually so reliable. However, observing the signs of poor configuration will create cognitive dissonance; a cerebral discomfort that tells you something is not quite right.

Margaret shared how she had learned to acknowledge her distress signals, defocus, and complete a self-assessment. She writes an entry in her "Notebook of Woe" to locate the source of cognitive dissonance that she is experiencing. This notebook is a tool to ask herself a series of questions. How do I feel about my work? What deviations am I experiencing from my model?

I love the idea of this notebook, and the message that Margaret delivered alongside it. "Failure provides an opportunity for learning and growth, but it's not comfortable and it's not easy". This constant and deliberate self-assessment is a method for continuous personal improvement, capturing our discomfort before failure has an opportunity to take root.

Bugs in Brains

I ate lunch with Anna Royzman and Aaron Hodder one day. Our conversation meandered through many testing topics, then Anna said something that really struck me.

"I keep finding bugs in people's brains, where do I report those?"

She was speaking about the way in which we, as testers, start to learn how to interrogate a piece of software based on the developer who coded it. When we've worked in a team for a long time, our heuristics may become incredibly specialised.
Aaron concurred and provided an example. In a previous workplace, he knew that the root cause analysis would lead in very different directions dependent on the developer that had introduced a bug. One developer would usually introduce bugs that were resolved by a configuration change for an edge case scenario. Another developer would usually introduce bugs that were resolved by a complete functional rewrite for a core system flow.

This was something that I could also relate to, but had never considered as anything unique. I'm now thinking more about whether testers should raise these underlying differences between developers and how we might do so.

Talking to Developers

Keith Klain gave a keynote address in which he spoke about the ways to successfully speak to management about testing. I found his advice just as applicable to the interactions between testers and developers.

Enter conversations with an outcome in mind. Manage your own expectations by questioning whether the outcome you are seeking is a reasonable one. There is a common misconception that testers are wasting developer's time. Having one specific goal for every interaction is likely to keep your conversations focused, succinct and valuable, which will help to build rapport.

Know your audience and target your terminology accordingly. Even if you don't have the same skills that they have, you can still learn their language; interactional expertise. You can talk like a developer without actually being able to do development. For example, learn what third party libraries are used by the developers, and for what purpose, so that you can question their function as a tester.

Work to build credibility with people who matter. If you don't join a team with instant status, remember that this can be built by both direct and indirect means. Cultivating a good reputation with people in other roles may help create respect with those developers who are initially dismissive of you as a tester.

* By the way, Margaret has an excellent article on this topic in the upcoming October edition of Testing Trapeze.

No comments:

Post a Comment