Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Dual N-Back Training for Intelligence

soakyourhead.com

Post reply on HN
Print view View original

31–32 of 32 posts

Re: Dual N-Back Training for Intelligence

#31

Post by gwern »

tocomment wrote: post #16

In the paper didthe difficulty increase automatically? How did that work?

It did increase automatically - the adaptiveness is a major part of the novelty - and it worked basically promoting you when you scored high (like >=80%) and likewise demoting you an N level. If you're curious, Brain Workshop has a Jaeggi mode where it uses the same scoring and promotion criteria.

Re: Dual N-Back Training for Intelligence

#32

Post by gwern »

emef wrote: post #4

"In the study performed by Susanne M. Jaeggi et. al., study subjects practiced the dual n-back test for about 20 minutes a day for up to 19 days. At the end of that time, the subjects were shown to have increased intelligences (as measured by gF)." An improvement in intelligence that becomes apparent after only 19 days smells too good to be true. Correct me if I'm wrong, but surely someone that practices IQ test prob…

Yes, that's one of the main criticisms - that the n-back task is too similar to the Ravens/BOMAT used to measure _Gf_.

See http://www.gwern.net/N-back%20FAQ#criticism and this paper I haven't incorporated yet which reviews & criticizes a lot of the pro-WM training literature: http://psychology.gatech.edu/renglelab/publications/2010/shi...

Post reply on HN