DiagnosticsHow The Carrot Approach Facilitates Learning
People who are rewarded for making correct decisions learn quickly. While
the "carrot" approach may produce favourable results, little is
understood about how rewards facilitate the learning process. Now, in a
paper published this week in the online open-access journal PLoS Biology,
team headed by Dr. Burkhard Pleger of the Max Planck Institute for Human
Cognitive and Brain Sciences, Leipzig, and the University College London
have
demonstrated that the "reward effect" not only supports the improvement of
higher cognitive abilities, but also how brain function in the cortex
can be enhanced. Intriguingly, they see that the reward effect can be
strengthened using dopaminergic compounds. Targeted manipulation of
dopamine
levels, thereby enhancing the "teaching signal" in the brain, could open
up new possibilities in the treatment of patients, for example, after a
stroke.
Previous work has shown that if a decision leads to a successful outcome,
it is registered in the brain"s reward system. The reward stimulus is then
relayed to the area of the brain which was responsible for making the
decision. In this way, the brain optimises its processes for improved
performance each time. "It was not known until now, however, whether this
mechanism also applied to functions of the somatosensory cortex, which
process the skin"s sense of touch, for instance," explains Dr. Pleger.
To answer this question, the researchers designed a "game" for their
subjects. Electrodes were attached to both index fingers of the subjects.
In
each trial, two electric currents, each with a different frequency, were
successively applied to subjects" fingers via the electrodes. Subjects had
to decide whether the first or second electric current had a higher
frequency. If they were correct, a monetary reward was displayed on a
screen. The
fact that the reward effect works when it is displayed visually had
already been confirmed in an earlier study. The amount of the reward was
varied
from trial to trial. The result: depending on the size of the reward, the
subjects were able to subsequently make the correct decision with improved
accuracy. "As well as the effects on higher cognitive processes which were
already known, it shows that the reward effect also influences
somatosensory processes," says Dr. Pleger. "It turns out to be stronger,
the higher the reward."
In addition to this, the researchers were interested in the role of the
neural transmitter dopamine. Therefore, subjects were divided into three
groups before the experiment began. The researchers administered the
dopaminergic compound Levopoda to the first group, and the second group
received
the dopamine inhibitor Haloperidol. The third group were treated with a
placebo as a control. The effect was clear-cut: The effect of the reward
was
greatest in group whose dopamine levels had been raised by Levopoda.
Subjects in the placebo group also learned with each trial, albeit by
less. The
reward effect was totally absent in the group of subjects who received the
dopamine inhibitor.
"Apparently, the interaction between the regions of the reward system and
the somatosensory cortex are mediated by the transmitter dopamine," says
Pleger. This discovery opens up interesting possible uses for medicine.
The researchers write that in the future, targeted use of dopaminergic
compounds could be used to aid the rehabilitation of stroke patients, for
example. In theory, applications for pharmaceutical "learning boosters"
are also conceivable. However, caution is needed in this. "A raised
dopamine level in the brain has already been identified as the cause of
mental
illnesses such as schizophrenia," warns Dr. Pleger, "so too much is also
not good, and can even be dangerous."
Funding:
This work was supported by Wellcome Trust programme grants (to
RJD and JD) plus the Medical Research Council UK (to FB, JD, and CCR) and
the
German BMBF (to FB). JD held a Royal Society Leverhulme Trust Senior
Research Fellowship. The funders had no role in study design, data
collection and
analysis, decision to publish, or preparation of the manuscript.
Competing interests statement:
The authors declare that no competing
interests exist.
Citation:
"Influence of Dopaminergically Mediated Reward on Somatosensory Decision-Making."
Pleger B, Ruff CC, Blankenburg F, Klē¶ppel S, Driver J, et al. (2009)
PLoS Biol 7(7): e1000164. doi:10.1371/journal.pbio.1000164
PLoS Biology