Wednesday, August 3, 2011

In praise of mistakes

We learn from mistakes. We learn from failures.

No, that's not quite it. We learn from variation. We learn from looking for success, and then replicating that success.

Think of how evolution works. Variation and the more successful variants are more likely to survive. There are lots and lots of mistakes. Lots of variations that are no better, and in most cases are worse, than the current best. But now and then, a variation turns out to perform better than the current best. And that's the one to build the next generation on.

So, why praise mistakes? Mistakes scare us. Mistakes are generally considered failures and failures are considered bad. But that's just short sighted. We must have variations, experiments, in order to do better. And if we're afraid of mistakes, we're afraid of variations. If we're afraid of variations, we'll be too timid to improve.

We must embrace mistakes. We must embrace failures. But that is only half the story.

Evolution wouldn't work without death. Death is a strict evaluation function that is a judgement on whether or not a variation is an improvement.

When we're experimenting, when we're creating variations, it is just playing around unless we're keeping score. We must have an evaluation function that tells us whether a variation is better or worse than the current best. We must have an evaluation function that tells us to kill one variant and replicate another.

We must measure. We must keep track. We must evaluate.

So, what do we need to improve education? We need experiments. We need to measure their effectiveness on student achievement. We need to tolerate mistakes. And we need to replicate the variants that give better results.

That's the essence of continuous improvement.

Posted at 12:00 AM (permalink) 4 Comments View/Leave Comment
Keywords: continuous improvement