The data in this repository were collected from four marmosets performing a serial reversal task in a touchscreen apparatus. In each session, marmosets were presented with the same two visually distinct stimuli and had to determine, through trial and error, which stimulus to touch in order to earn a reward ("baseline discrimination phase"). Once criterion was reached in this initial phase, the stimulus-outcome contingencies reversed, such that responses to the previously unrewarded stimulus were now rewarded and vice versa (“reversal phase”). Trained marmosets were implanted with indwelling cannula in both the caudate nucleus and putamen. Varying doses of the GABA-A agonist muscimol were infused in experimental sessions to inactivate these areas separately in independent sessions. Performance following infusions of muscimol, as measured by errors committed (Figure 3) and trials completed, was compared both against control sessions with no infusions and sessions in which saline vehicle was infused into one of the target areas. Response latencies were also assessed and strategy usage was evaluated using a win-stay / lose-shift analysis. Performance across three timepoints spanning the length of the infusion experiment was analyzed to determine whether there were any effects of time in experiment, cannulation, or number of infusions.

This repository contains data supporting the analysis and figures reported in the article "Selective Role of the Putamen in Serial Reversal Learning in the Marmoset".