Tag Archives: continuous improvement

Who kept the dogs in? – Continuous Improvement Part 2 of 2

In part 1 of this post I was talking about how, when continuous improvement does work, it always strikes me as a much less stressful way to work and I was wondering why that is.

Although there may be conceptual parallels between Thorndike and AGILE, I don’t feel I have yet gotten to why continuous improvement feels like a more forgiving and easy environment. I was thinking maybe the work of Martin Seligman can shed some light on this.  Seligman (and Steve Maier) put dogs into situations where there was nothing they could do to avoid painful electric shocks. They found the dogs behaved in much the same way as depressed people. Other dogs could stop the shocks by pressing a lever. Still other dogs were not shocked at all. After this, Seligman and Maier put the same dogs in boxes (there they go again with their animals and boxes) and experimented with making them shuttle from one side of the box to the other. He motivated them to move using electric shocks. If they jumped over a small partition they escaped the electric shocks. I find this pretty grim stuff and am not at all clear this was worth it. Still, it is interesting to note that the dogs who had not been shocked and the dogs who had been able to do something about it were able to escape by jumping over the partition. The dogs who had been helpless sadly laid down and were shocked. They did not try to escape.

I am wondering if learned helplessness goes some way to explaining why continuous improvement environments are less stressful. We always feel we can try something new.

There is an interesting take on learned helplessness in people here


Who let the cats out? – Continuous Improvement Part 1 of 2

Sometimes I work in agile environments where work goes in cycles (sprints or timeboxes) and at the end of each cycle we ask “what can we do better next time?”. This is hard to get going in highly political environments or teams that are a mixture of client and supplier staff (people don’t like to rock the boat or accept liability) but where it has worked well, I was struck by how less stressful work became. We weren’t counting on things being right, only better than before.

It brought to mind the work of Edward Thorndike. From what I understand he spent much of his time shutting cats in boxes. You might be thinking he should get out more and funnily enough, that is exactly what he wanted the cats to do. Thorndike’s cats tried any number of motions, at random before triggering a lever that operated the door to open. Having stumbled upon how to operate the door, the cats would be put back in the box (doesn’t seem right somehow) and then take progressively less time to operate the lever to open the door. The mental mechanism he identified is called operant conditioning.  He contributed greatly to the behavioural branch of psychology with his systematic studies on learning. I really like the graph that shows what must have been one of the first “learning curves”. Reminds me of a burndown chart. Likewise, Thorndike’s operant conditioning makes me think of an agile environment with it’s “try things out to see if you hit on the answer” style of continuous improvement.