Video transcript
What’s better for improving skills based performance (and remember that language learning is skills based performance). Punishment or praise?
If you want to know the surprising reason for the answer you may well have chosen, keep watching.
I’m Jo gakonga from ELT-Training.com. I’ve been a teacher for over 30 years and a teacher educator for over 20 of those, but this was a new thought for me and I wanted to share it with you.
I’m reading a great book at the moment by a Nobel prize winning psychologist and this little nugget caught my eye as relevant to language teachers.
Let’s start with this conundrum:
There is lots of research evidence in training skills based on pigeons, rats, humans and other animals that rewards for good performance work more effectively than punishment for mistakes… so if you lean in that direction, great!
The surprising fact, though, is that when someone does something well, even if you praise them, they’re quite likely to do it WORSE the next time and if they do something badly and you chastise them, they’re quite likely to do it BETTER next time.
How can these two things BOTH be true?
Apparently, it’s all down to the fact that our performance varies statistically and there’s a phenomena called Regression to the Mean. Let’s take irregular past tense verbs as an example. You might gradually remember to use them over time , but it’s going to be a two steps forward, one step back kind of journey before you master them. This means that if you do a particularly accurate job on one occasion, it’s statistically quite likely that you won’t be quite so successful the next (whatever your teacher does) and if you have a particularly bad day and completely forget them (whether or not your teacher shouts at you) you’re likely to do a better job next time.
Interesting, don’t you think?
So the upshot of this is that praise IS a better way forward- just don’t expect immediate results! What do you think? Have you noticed anything like this in your classroom? Let me know in the comments below.
