Dissociation between learning from one’s own mistakes and from errors made by others in Parkinson’s patients

In everyday life, one constantly faces situations where one has to make choices. These decisions are guided both by consequences, rewards and punishments, of one’s own choices in similar situations in the past, as well as by having observed others’ choices getting rewarded or punished. It has been, however, poorly known whether learning by observing others and learning from one’s own mistakes rely on the same vs. different underlying neural mechanisms. 

In Parkinson’s disease there is a bias to learn more from negative than positive feedback, which has been presumed to be due to loss of striatal dopaminergic function in the disease. It has not been, however, investigated whether there is a similar bias with respect to learning by observing successful and erroneous choices of others. Dissociation between these two types of learning in Parkison’s disease would suggest that dopaminergic system plays a differential role in trial-and-error vs. observational learning.

In their recent study, Kobza et al. (2012) investigated this highly interesting question in a total of 19 Parkinson’s patients and 40 healthy controls, divided into separate groups that were exposed to highly similar trial-and-error and observational learning tasks, respectively. The results showed that while Parkinson patients (who were off medication) exhibited a typical bias to learn better from negative than positive feedback when they were actively performing the task themselves, those Parkinson patients who learned by observation showed similar pattern of results as the healthy control subjects who learned better from positive than negative feedback. 

These findings indicate that there is a dissociation in the involvement of phasic dopamine activity in trial-and-error vs. observational learning. Dopaminergic activity, while clearly implicated when learning by getting positive and negative feedback based on one’s own active behavioral choices, would seem to be little involved in observational learning. These findings clearly justify further studies into the neural mechanisms underlying the ability to learn by observation.

Reference: Kobza S, Ferrea S, Schnitzler A, Pollok B, Südmeyer M, Bellebaum C Dissociation between active and observational learning from positive and negative feedback in Parkinsonism. PLoS ONE (2012) 7: e50250. http://dx.doi.org/10.1371/journal.pone.0050250