Thoughts on ‘Learning from failure or success’
Several weeks ago Mark Gould blogged about whether we learned better from instances of failure or instances of success. The post is interesting and I suggest you give it a read.
I don’t have any specific comments on what Mark has written beyond saying that I mostly agree with what he wrote; instead I would like to share some thoughts stemming from my own experience and views.
Bottom line, I’m of the view that failure is a better source of learning than success. Here’s an example which typifies my position.
Mrs. Knowledgethoughts and I are expecting our first child. As part of the process, every old wives’ tale and theory comes out of the woodwork. One of my favourites is that the severeness of morning sickness is indicative of the baby’s sex. The thinking goes that if there is morning sickness the child is more likely to be a girl, if there isn’t, a boy. A quick squiz on various baby forums shows this theory to be false and also shows cognitive fallacy at play.
“It’s true for me, so it must be true”
Which works for about two posts until contrary evidence is presented.
Correlation does not imply causation, and learning only from success often stems from correlation, the same type of learning that led us into the credit crisis (these CDSs make money now so they must do so in the future). It is just as easy to find examples of failure should you only look so far, and a single instance of failure proves the theory incorrect. Success can attributed to any number of variables, and that attribution may or may not be correct. Failure forces us to confront our Cognitive Bias, our tendency to make errors in judgement based upon cognitive factors. Failure forces an adjustment of our mental model.
Our adjustments won’t necessarily be correct, but in the process we have gained two valuable pieces of knowledge:
- our mental model is not correct… yet; and
- our understanding of the situation is not perfect
The first is obviously important, because you are required to either alter your view or bury your head in the sand. Should you chose the former, you will be better off in the future. Lawyers engage in this type of thinking whenever they construct an argument. Rather than accept raw ideas, they examine ideas for flaws, seeing how they stand up to changes in circumstance.
The second point is even more important because it shocks us into the unavoidable conclusion that our belief was mistaken and that future beliefs may also be incorrect and will require more rigorous conception.
I’ll close this post by sharing an example from my programming past:
When writing software, my first pass at coding is full of holes and flaws. I usually focus on implementing the perfect scenario. The user types the right thing, hits the right button and gets the right answer.
Perfect, but all I really know is that my software can handle correct input. I could run the software 100 times and get a right answer, but I haven’t learned anything about what’s wrong with it.
Fortunately I know from hundreds of previous bugs that my limited success is insufficient, and set about seeing how my system copes with unexpected inputs; users entering dates in the wrong format, missing out fields or clicking “submit” 50 times. The system will then fail time after time after time, whereby I refine the code (the model) and over time it becomes more robust, a process only becomes possible by intentionally seeking failure.
Ever come across a piece of software that didn’t work?
Don’t you wish the developer responsible had failed just one more time?
