Wednesday, June 13, 2007

Tinker Toy Traffic

This is your working memory:

This is your working memory on facts:

You'll notice that you only have 5-8 slots (most recent research says 5-7, but my Tinker Toy box doesn't include a 7-hole connector, so let's just pretend) in working memory, into which you can jam information. If you try to think about a new piece of information, when your working memory is already full, a random piece of the information that you had originally been thinking about will be shoved out of working memory, in order to make room:

When you try to keep track of too many bits of new information at once, you end up with a kind of traffic jam, where you are constantly swapping bits of information in and out of your working memory. It makes your brain hurt. It can make you tired and cranky. And it leads to what most of us think of as "careless errors" -- where you lose track of information you really do know, and you end up with a mistake in your work, even though you tried really hard not to make a mistake. But you were NOT careless, and you are NOT stupid. What actually happened is that you suffered cognitive overload, by trying to keep too much random information in your head, at the same time. It simply cannot be done.

So, how is it that experts (and expert learners) never seem to make careless errors? How is it that they always seem to "just see" the answers that they need, without having to think so %^#%@ hard? Do they have more slots in their working memory? (Nope.) Are they better at swapping information in and out of their slots? (Nope.) Faster at it? (Nope.)

The secret is in the size of their "schemas."

Expert learners never simply try to remember all the bits of information they are taught. They constantly ask themselves, "How is this idea similar to something -- anything -- that I have seen before? How does it fit in with other things I know? What patterns can I find in the new information? What familiar memories can I use to help me make sense of this?

In asking and answering these questions for themselves, they build what are known as "schemas": chunks of related information, which travel together...

and conveniently fit into a single slot in working memory!

This allows them to have much more information in working memory at one time, and so makes it much easier to think about things (new and old), and to find correct answers, without the headache of trying so hard.

written by Larisa Naples, M.S.Ed., Ph.D.

Email, bookmark, or share this post:

* Photos in this post are protected by a creative commons attribution license: all photos by Larisa Naples

1 comment:

Raf said...

Hello Larisa,

A very nice article. Amazingly, this is the type of approach I use in the relationship networks algorithms I helped develop at my company, SeeqPod (in Emeryville, CA). I look forward to more of your writing and hope to hear from you. I'd love to give you more details about my work, chat and catch up.


Raf (from RPI)