Practicing most of the games that are available on Lumosity will
make you better at that game. Many of them will make you better at
very similar games. However, only one of their games has been shown
to improve performance on dissimilar tasks; that game is Memory
Lane, an implementation of the dual n-back task. Dual n-back tasks
are available from many other sources, many of which are free. From
what I've heard from people who've tried both commercial DNB
programs (like High IQ Pro) and the most popular free one
(
http://brainworkshop.net), most prefer the free one because of its
rich feature set.
Of course, being that I'm one of the developers for Brain Workshop,
you should assume I'm biased.
The article published in the Mensa journal that Pontus mentioned was
disappointing to anyone who actually read it (and not just the
marketing blog summary):
The problems appear with the results. Results were only significant
for the divided visual attention task, FSWMT and RSWMT. The letter
memory task results were not significant. Looking at the between
groups analysis the results were only significant for the divided
attention task and the FSWMT block test. So, the difference between
the active and control group was only significant for 2 of the 4
tests.
Now, consider those 2 tests. The divided visual attention assessment
is described as being almost exactly the same thing as the
birdwatching task that the subjects were trained on. And, the FSWMT
assessment is really quite similar to the Monster Garden task. So,
training subjects may have gained simply because they built skills
specific to the birdwatching task and FSWMT.
http://groups.google.com/group/brain-training/browse_thread/thread/fdbd0af92c1dad37?tvc=2&pli=1
Lumosity's games are fun, varied, and well put-together. However,
they are not particularly effective. If you want something that
really works, I suggest you stick to n-back tasks until something
else reaches the market that has actual science to back it up.
Jonathan