Delayed Reinforcement Choice Task Protocol

The task was based on delayed reinforcement choice tasks that have been described before [73 (link),74 (link)]. The session began in darkness with the levers retracted; this was designated the intertrial state. Trials began at 40-s intervals; the format of a single trial is shown in Figure 2. Each trial began with the illumination of the houselight and the traylight. The rat was required to make a nosepoke response, ensuring that it was centrally located at the start of the trial (latency to poke was designated the initiation latency). If the rat did not respond within 10 s of the start of the trial, the operant chamber was reset to the intertrial state until the next trial began and the trial was scored as an omission. If the rat was already nosepoking when the trial began, the next stage followed immediately. Upon a successful nosepoke, the traylight was extinguished and one or both levers were extended. One lever was designated the Large/Uncertain lever, the other the Small/Certain lever (counterbalanced left/right). The latency to choose a lever was recorded. (If the rat did not respond within 10 s of lever presentation, the chamber was reset to the intertrial state until the next trial and the trial was scored as an omission.) When a lever was chosen, both levers were retracted and the houselight was switched off. Choice of the Small lever caused the certain delivery of one pellet; choice of the Large lever caused the delivery of 4 pellets with a particular probability (see below). When reinforcement was delivered, the traylight was switched on. Multiple pellets were delivered 0.5 s apart. If the rat collected the pellets before the next trial began, then the traylight was switched off and the time from delivery of the first pellet until a nosepoke occurred was recorded as the collection latency. If the rat did not collect the food within 10 s of its delivery, the operant chamber entered the intertrial state, though collection latencies were still recorded up to the start of the next trial. The chamber was then in the intertrial state and remained so until the next trial. There was no mechanism to remove uneaten pellets, but failure to collect the reward was an extremely rare event. The large-reinforcer probability was varied systematically across the session as follows. A session consisted of 5 blocks, each comprising 16 trials in which only one lever was presented (8 trials for each lever, randomized in pairs) followed by 10 free-choice trials. The probability that the large reinforcer was delivered, given that the Large lever had been chosen (p_reinforcer), varied across blocks: it was initially 1, 0.5, 0.25, 0.125, and 0.0625, respectively, for each block. As trials began every 40 s and there were 130 trials per session, the total session length was ~87 minutes; subjects received one session per day. Choice ratios (percentage choice of the large reinforcer, for each trial block) were calculated using only choice trials on which the subject responded.