The basic procedure was similar to that of earlier experiments on the differential reinforcement of two classes of discriminated interresponse times (e.g., Hawkes & Shimp, 1974). Two temporal patterns, a shorter and a longer, were concurrently reinforced. A variable-interval schedule arranged a distribution of minimum interreinforcement intervals and a pattern-selection procedure randomly assigned each reinforcer arranged by the variable-interval schedule to one of the two patterns.
Consider the details of the variable-interval schedule. A single timer queried a random number table every one sac to determine whether to arrange a reinforcer. The probability of arranging a reinforcer every one sac was 0.050, so that the schedule was a random interval 20 sac schedule. Once a reinforcer was arranged, the 1-sec timer stopped until the reinforcer was collected, or until the reinforcer was cancelled by an excessively long pause. If a rat paused for longer than the upper bound of the longer reinforced pattern, an arranged reinforcer was cancelled. This was to discourage long pauses.
Now consider the pattern-selection procedure. The durations of the reinforced pairs of patterns, shorter and longer, are shown for each condition in Table 13-1. Throughout the experiment the longer pattern either was exactly or approximately three times longer, and also three times wider, than the shorter pattern. Thus, as the absolute duration was varied, relative duration remained fixed. The assignment of each reinforcer arranged by the variable-interval schedule to a pattern was random in the sense that the patterns were equally likely to be selected and the pattern chosen for one reinforcer was independent of the pattern chosen for the preceding reinforcer.
Reinforced classes of interresponse times|
(lower and upper bounds in sec)
|1||1.5 - 2.5||4.5 - 7.5|
|2||2.5 - 4.0||7.5 - 12.0|
|3||4.0 - 7.0||12.0 - 18.0|
|4||1.5 - 2.5||4.5 - 7.5|
|5||1.0 - 2.0||3.0 - 6.0|
|6||0.75 - 1.5||2.0 - 3.5|
|7||0.5 - 1.0||1.5 - 3.0|