The delay-discounting task

The delay-discounting task we use is adapted from that first designed by Evenden and Ryan (1996). Each session typically lasts 100 minutes and consists of five blocks of 12 trials, each lasting 100 seconds. Each block begins with a pair of forced choice trials which consists of one presentation of the left lever and one of the right in a random order. Throughout the task, a response on one lever produces a reward of one pellet (lever A) whereas a response on the other produces a reward of four pellets (lever B). The position of these levers (left or right) is kept constant for each rat, but is counterbalanced between rats. The delay between responding on lever A and the concomitant delivery of reward (dA) is always 0s whereas the delay between responding on lever B and the delivery of reward (dB) increases within the session in a step-wise manner between blocks from 0 seconds in block 1, to 10s in block 2, 20s in block 3, 40s in block 4 and 60s in block 5.

A new trial begins every 100s with the onset of the houselight and traylight. The rat then has 10s to nosepoke in the magazine to trigger presentation of the two levers, and a further 10 s to make a lever-press response. Failure to respond in either 10 second period results in the trial being recorded as an omission and a return to the inter-trial-interval (ITI) state until the next trial is due to begin. Once the rat has responded on one of the levers, both levers retract and the houselight and traylight turn off. Food delivery, signaled by re-illumination of the traylight, occurs either immediately or after a delay. An inter-trial interval of variable length then follows depending on the choice made, so that each trial lasts 100 seconds. The length of each trial is kept constant so that the rate of delivery of reinforcement associated with both behavioural responses is identical, preventing any differences from influencing choice.

This task can also be run in the five-hole boxes using the far left and far right holes (holes 1 and 5 respectively) instead of levers A and B. We have also recently modified the task to allow for intra-cerebral infusions of drug. In this version, the number of blocks is reduced from five to four, and the corresponding delays to the large reward are condensed (0, 15, 30, 45s). The total trial length is also reduced to 70s and the session length to 56 mins.

Evenden JL, Ryan CN (1996) The pharmacology of impulsive behaviour in rats: The effects of drugs on response choice with varying delays of reinforcement. Psychopharmacology 128(2): 161-70