adjusted amount of the immediate side gives an estimate of the subjective valuation of the delayed reward by the end of the session. The adjusted amount of the immediate side was limited to 2 seconds so that it never became larger than the amount on the delayed side, as experience with very large immediate rewards might interfere with mice being able to reliably assess reward magnitude. This was a departure from the Richards and colleagues (1997) paper, which also included a side bias term in the numerator. As we capped the adjusted amount at 2 seconds (the size of the delayed reward), we were not sensitive to biases toward the delayed lever, so we excluded the bias term. Therefore, the range of adjustment was always between 0 and 2 seconds sipper access time. To expose the mice regularly to both immediate and delayed reinforcers, a forced trial on the opposite lever was instituted after 2 consecutive selections of the same lever. In a forced trial, only 1 lever was active, and only that light came on after a nosepoke. In addition, there was no adjustment of the immediate reinforcer resulting from forced trials.
