Logic Trees
There is a simple way of diagnosing a problem, and then finding a solution. It is called “the five whys?”
No doubt you have heard of it, it is a way of refining a problem until you get to its cause.
When faced with a problem ask why it happened. Then ask again. Repeat until you get to the point where you understand what is going on and can do something about it.
Here is a nice simple example, imagine you were late to work and had to explain yourself to the boss. The conversation could run a little like this:
As an aside, in my experience claiming your child was ill is a far better excuse than admitting you had a night out on the tiles
There are a couple of problems with the five whys
- If you are on the receiving end it is remarkably frustrating (I’ve just *!#*@!? told him why!)
- It assumes that all cause and effect chains are linear, there is only ever one answer to why.
The diagnostic tree
This is a more sophisticated approach. It realises that there can be more than one reason why something happened and allows you to capture them all.
As it flushes out different possibilities (not just one) it triggers a host more questions:
- Did you forget to switch on the alarm clock because you had a late night?
- Which was the biggest problem, getting up late or the traffic jam?
A note of caution: Do not take this approach when explaining your tardiness to the boss (stick with the sick child).
According to Arnaud Chevallier (writer of an excellent problem solving blog) there are four rules when creating a diagnostic tree:
- Repeatedly ask “why?” as you move from one level to the next (just like the five whys).
- Start with open questions and move towards more concrete answers as you move down through the levels.
- Make sure the branches are mutually exclusive (don’t overlap) and collectively exhaustive (don’t miss anything). This is tricky but valuable, here is an example here.
- The answers should be insightful. If an answer fails the “so what?“ question it isn’t insightful. (My daughter once phoned me when she was with her mother coming to pick me up from work. I asked her where she was, the answer that came back was “in the car”…)
The solution tree
Another reason to use logic trees is that they work both ways.
If you ask “how?” rather than “why?” you will flush out a whole host of solutions. Here is an example (click on it if you’d like to see a larger version).
The perfect tree doesn’t exist
You could order the branches in any way you like, and argue forever about the categories. That would miss one of the really valuable points of using a tree, which is…
Communication
Trees are a powerful way of structuring communication. They show how you are thinking about a problem.
- Are the branches really mutually exclusive?
- Is there another solution or cause that you have missed?
- Do the answers pass the “so what” question?
- Could you measure the number of times an outcome happens and overlay some data?
If I show you my thinking no doubt you can add to it and I will get a better answer.
A final thought
It is important that the branches are mutually exclusive but the solutions don’t have to be.
Perhaps the way to make sure you get to work on time is to find a new job closer to your house and use the pay rise to buy a Porsche.
Do you really want to work for somebody who is such a nit picking time keeper anyway?
If you enjoyed this post click here to get the next
Read another opinion
Image by Martijn de Valk
Leave a Reply