Far transfer effects of Cogmed confirmed using machine learning techniques | New article in Nature

Anyone following the field of cognitive training will be well aware of the ongoing debate about whether programs such as Cogmed can lead to generalizable improvements of capacity (far transfer) or if the effects are merely in the form of strategy improvements on tasks similar to those trained (near transfer).  With the publication of Julia Ericson and Torkel Klingberg’s new article in Nature, “A dual-process model for cognitive training”, we hope this question can now be considered firmly answered.

In the article, Ericson and Klingberg demonstrate a method to distinguish between the two processes (capacity and strategy), solely by studying the training curve of an individual trainee. Analyzing training data from 1300 children using machine learning algorithms, they found that strategy improvements occur on average the first three days of training and account for 44 percent of the improvement on the trained task. The rest of the improvement, 56 percent, was shown to be capacity improvement (far transfer) which occurs gradually over the course of the entire training period. 

These findings highlight the effectiveness of Cogmed, demonstrating that the benefits extend far beyond initial strategy enhancement. They underscore that sustained, regular training leads to lasting increases in cognitive capacity.

So there you have it: Cogmed works.

Read the full article here:

Cogmed Team

Assembling neurons since 2002






published articles

"In our practice, we saw student after student go through the Cogmed program and find that they could now better manage their lives."

"I started feeling the benefits pretty quickly. I was staying engaged and focused for much longer in class, my memory was getting sharper, and that kind of brain fog that comes with a concussion was clearing up."

"Coming across this program was really a blessing, because it's hard to find
something that's as scientifically based as Cogmed is."