Working Memory

Richard Lanyon rlanyon at plato.wadham.ox.ac.uk
Tue Aug 31 08:33:44 EST 1999


On Tue, 31 Aug 1999 21:12:37 +1000, John wrote:
> Can anyone provide me with a reference for the research re those magic
> numbers 7 +-2 for working memory capacity? I have a vague memory of a name:
> Mountcastle.

I thought it was Miller. I have a reference here in a photocopy of
mine which says "Miller and Madow" though I'm missing the bibliography
that goes with it so that doesn't help much. Incidentally, the 7 +/- 2
stuff comes from early information theory analysis. Information theory
is often badly implemented, as there's a subtle problem whereby small
sample sizes systematically bias (upwards) estimates of information
transmission. There are many bias correction techniques, and some of
them are a lot better than others.


-- 
Richard
for the self-assured
I have no cure




More information about the Neur-sci mailing list