Tuesday, August 04, 2009

Conventional Wisdom

We've all hear the old saw about how we learn more from our mistakes than our successes. Of course this doesn't explain the Bush family nor why the Democrats can't seem to profit from GOP miscues. But it does make a recent study published in the July 30 issue of the journal Neuron very interesting.

The article is titled Learning Substrates in the Primate Prefrontal Cortex and Striatum: Sustained Activity Related to Successful Actions And according to this article, from Earl K. Miller's lab at MIT, single behavioral outcomes influence future neural activity and behavior. Later behavioral responses are more often correct and single neurons will more accurately discriminate between the possible responses when the previous response was correct.

Neat stuff, huh? Their research indicates that we only receive positive feedback from being right. Being wrong doesn't change the subsequent choices we make. Of course this research was conducted with monkeys, not people. But that brings me back to George W---we have a good amount of data showing that he didn't learn from his mistakes either. I'm just not sure which species of subject in which to place him. Just joking, I think.

Actually research has been done on humans along these lines. There was a study published last year that indicated that young children below the age of 12 didn't learn very well from mistakes but that changed during their teens. (where they found teens that learned from their mistakes wasn't revealed unfortunately---I can think of many parents that would love to know) Young children, like the monkeys in the MIT study, only showed strongly positive learning when praised for being right. Feel free to make conclusions from that if you want to.

That second study was from the 17 September 2008 issue of The Journal of Neuroscience, if you want to read it. Evaluating the Negative or Valuing the Positive? Neural Mechanisms Supporting Feedback-Based Learning across Development


Teresa said...

The problem with the thinking is that negative feedback doesn't seem to be a factor in this study. Positive feedback is conducive to learning because there's a payback. That's not surprising to anyone. It very well might make someone do the same thing if presented with the chance. (Think... getting your allowance if you keep your toys put away) But if not learning something has no consequence either way, it may not be important to learn it. And you can learn things from negative response too, such as if you play with fire, you will get burned. You will probably stop doing that fairly quickly without anyone praising you for remembering not to do it!

The real problem with this is that some people are susceptible to one or the other of these reinforcements. If no one is there applauding them, it's not worth learning it... or unless you there is a punishment for NOT doing something, it's not worth learning it (in other words, we take our learning cues from what others think about the activity's worth, rather than learning for the merit itself).

Though I'm sure that there was encouragement along the way, most of my learning hasn't been a result of feedback. (Other than grades, that is... and they were never important to me). I would have thrown these studies way off, I fear. (or maybe not fear...)

GPV said...

Hi Ut,
Nice prospecting about people being wrong-no positive feedback.
This is plain logic: If the guy is wrong about anything that's because he's been mislead by wrong informations about that thing and if he didn't learn from that it means that he's stupid,now how can you make someone understand that he's wrong,knowing that he won't know that he's stupid...........
Hard task,don't you think so?

tiff said...

What happens if you're never wrong, like me?


rosemary said...

sometimes you dazzle me with your brilliance. I must be related to tiff because I am always right too.

simplycol said...

Don't know about the studies on kids and teenagers... but with dogs positive reinforcement and feedback works wonders. I have heard that a dog equates somewhat to a two year old child.. not sure where the monkey would fit. :-)