Lets Test Session: Alan Richardson - Testing Hypnotically

Posted By Duncs ~ 2nd July 2012

Alan uses his 20+ years experience of hypnosis to influence his approach to testing. In his session, Alan aimed to give us a taste of how hypnosis can be used to help with our testing. Slides for the Testing Hypnotically Session can be found here

Alan Richardson in hypnotist pose

Alan set up the session by showing us his MoRim model of software testing

  • Model
  • Observe
  • Reflect
  • Intent
  • Manipulate

 

To help us make the leap between hypnosis & testing during the session, Alan gave us some terminology (not a metaphor, analogy or simile!):

  • When I say Hypnotist, hear tester
  • When I say Hypnosis, hear testing
  • When I say trance, hear process
  • When I say patient, hear system

 

Alan got us thinking from the outset with some statements which seemed to contradict each other:

“You already know hypnosis”

and

“There is no such thing as hypnosis”

and

“Psychotherapy is one of the arts of High Weirdness that we can apply to Software”

(a quick search for “high weirdness” didn’t prove directly fruitful, but it appears hypnosis, along with UFOs & drugs, lives under the term. Stuff that can’t obviously be explained maybe?)

 

We were then taken on a brief history of hypnosis:

 

Alan rounded off this introduction to hypnosis with an interesting quote:

“Work in context with the patient (system) you have and utilise what they give you”

This was tied in with automation - you can do hypnosis without knowledge of the specific patient in question via generic tapes & CDs, but the effect is limited. The comparison here is you can test without knowledge of the context, but it wont be as effective.

In the context of hypnotism, if you tried to apply the same approach with each patient/subject then you would be unsuccessful. Same with testing - if you try to apply a boiler plate approach to each project or feature you test, you are not going to be successfully testing it.

 

Alan then produced some tools of hypnosis (I hope your still mentally connecting to testing) & went through each one in detail:

  • Confidence
    (Own your own approach, trust your observations & model)
  • Language
    (“Language is the main weapon of a hypnotist, and it is ours too.”)
  • Ambiguity
    (“It’s simpler than that”)
  • Observation
    (physiological/verbal, subjective/objective, not/there)
  • Presupposition
    (make an assumption & challenge it)
  • Feedback (+ve/-ve which has highest value?)

So this is how I interpret this wise advice - be confident in your approach to testing & question if you’ve fully understood what you’re testing. If you appear unsure about testing approach, people are going to question its validity. Ensure you tackle the problem from different perspectives, considering both subjective & objective viewpoints. Think about what assumptions you may have made in order to start testing - how could you break them down & challenge them? Does it provide another angle to test from?

Finally, thinking about providing information about your testing to the project & stakeholders - how would they want that information to be delivered? How would you want the information to be delivered if you were in their shoes? Yes, sometimes we have to deliver bad news, but you don’t need to be a little tell-tale about it.

 

There are several hypnotic techniques relevant to testing:

  • Induction
    (putting subject into “hypnotic trance”, heighten expectations, focusing attention etc)
  • Post Hypnotic Suggestion
    (suggestions made to the subject after emerging from hypnosis can affect their behaviour)
  • Dissonance/Confusion(uncomfortable feeling caused by having conflicting ideas simultaneously. Used to question clients choices them deem as correct)
  • Pattern Interrupt
    (breaking a coherent flow of ideas or thoughts)
  • Future Pacing
    (ask subject to imagine themselves in the future, normally used to see if change process is successful)
  • Go there first
    (put yourself in the situation where you want your subject to go
  • Embedded Command
    (bypass the conscious & plant ideas in the subconscious)

Alan went through each of these techniques & how they can be applied to testing (see slides for high level comparison).

 

The end of the session focussed on effective communication:

  • You can not not communicate
  • All communication is suggestion
  • Communication is the message received
  • Communication is more than one way

If you swap out “communicate” for “test” it helps us to make the comparison.

You can not not test - everything you do & the order you do it in may impact the result of the test. The system is testing you.

All testing is suggestion - you provides input into the system but you don’t know what the system has done as result of your input. You need to check the output. The system suggests things to you as well

Testing is the message received - not the intent, medium or message. Does the message received reflect / match the message sent. What did the system communicate to you?

Testing is more than one way / multi-directional - You & the system are involved. Who else is involved? Remember yourself & others.

Alan left us with a question

“What will happen when you think of your management approach as effective communication?”

(relating to test plans, strategies, meetings, etc.)

I’ll try to answer that question!

If you were applying the model for effective communication, the documented approach will likely fall over because it is largely one directional & the message being communicated in the documentation may be unclear / ambiguous / not obvious to the reader. Yes, there are review stages & the like, but the time between original authoring, review feedback & updating the document is too great for it to be considered effective communication (i.e. its not real time, like in a meeting).

I see more dynamic management approaches with faster feedback loops as being more effective communication.

If we think our test artefacts are effective, we may be deceiving ourselves & not instilling faith in our ability to test.

This was a great session (& supported by an evening session Alan did) & has probably been the hardest for me to write up. Not sure why that is - maybe Alan has put me in a trance?