Sensitivity to pre- and post-reinforcer delays in self-control choice

• Post-reinforcer delay influenced rats' choices.
• Demonstrated the limitation of the matching law.
• General models of choice must consider all time components of choice alternative.

Rats were exposed to a concurrent-chains schedule in which reinforcer amounts, pre-reinforcer and post-reinforcer delays were varied between alternatives. Two independent variable-interval 60-s schedules were in effect during the initial links. Pre- and post-reinforcer delays were defined by fixed-time schedules. In one condition, rats chose between a larger reinforcer delayed by 0, 40, 80, or 120-s and an immediate smaller reinforcer. In the other condition, rats chose between an immediate larger reinforcer terminated with 0, 40, 80, or 120-s post-reinforcer delays and an immediate smaller reinforcer terminated with no post-reinforcer delay. In both conditions, choice was predicted from overall reinforcement density (i.e. reinforcer amount divided by total trial time). The results indicate that the preference for the larger reinforcer systematically decreased as both pre- and post-reinforcer delays increased. Although the effect of pre-reinforcer delay was greater than that of post-reinforcer delay, post-reinforcer delay itself influenced rats' choices in the present self-control situation. In conclusion, the present study demonstrated the limitation of local reinforcement density model of the matching law and all time components that constitute a choice alternative need to be weighted and incorporated into the general model of choice behavior.