From Kéri et al. Take Parkinson's Disease (PD) patients; give them a task where they must remember certain letters associated with pictures, but not other letters associated with pictures, in order to receive a reward. (There were also pictures with NO letters.) At baseline, the PD patients did the same as non-PD patients. After starting the PD patients on one of three dopamine agonists, they now remembered both kinds of letters (specified and distractor) better than non-PD (and non-medicated) patients. That is to say - after administration of dopamine agonists, they were better than non-medicated patients at remembering the rewarded stimuli as well as the distractors.
The core features of psychosis can be modeled as salience defects, and the working clinical hypothesis is that this is mediated by hyperactivity of dopamine in the mesolimbic system (which same feature, unfortunately but predictably, can also cause Parkinsonian symptoms). This is supported by the effectiveness of anti-dopaminergic anti-psychotics in treating psychosis. This paper is important in showing that control of salience is damaged by exogenous dopamine agonism.
Kéri S, Nagy H, Levy-Gigi E, Kelemen O. How attentional boost interacts with reward: the effect of dopaminergic medications in Parkinson's disease. European Journal of Neuroscience Online, 8 SEP 2013 | DOI: 10.1111/ejn.12350.