Reduction in sucrose reward magnitude without generalization decrement

Abstract

Groups of albino rats were trained for 24 days in a straight runway with one of two different liquid sucrose concentrations, 30% or 3%, as reward. Training trials were administered 2/day with an intertrial interval (lTI) of 3·5 min. This preshift phase was followed by an additional 24 days in which both groups received the 3% solution. A significant main effect of reward magnitude developed early in the preshift phase and persisted throughout the postshift phase. Since the number of daily trials and the ITI employed in this experiment have been shown in other designs to be optimal for the conditioning of the sensory consequences of responding with sucrose rewards, the data argue against a generalization decrement account of the successive contrast effect.

Successive reductions of liquid and solid sucrose rewards

“…The postshift conditions ended the preshift reward discrimination, unlike the earlier experiment, in which postshift trials were administered at 2 rather than 10 per day (Burns & Burns, 1978), but no clear SuNCE developed. The last tria1 of Day 1 in postshift could suggest contrast, but the lack of a Groups by Trials interaction coupled with (1) the results of a subsequent replication in which no crossover at all was observed (Burns, Note 1), and (2) the fact that the Trial 10 mean was unduly influenced by a single extreme score warrant the alternative interpretation.…”

“…In a pair of experiments examining sequential learning in paradigms other than the contrast paradigm, Burns (1976) showed that sucrose traces are clearly differentiated under a restricted set of conditioning parameters, two trials per day with a 3-to 5-min intertrial interval (ITI), a result that implies that the SuNCE would occur with sucrose under the proper training conditions. Yet a shift from 30% sucrose to 3% sucrose failed to produce the SuNCE following training under these optimal aftereffect conditioning parameters (Burns & Burns, 1978). The postshift procedure of that sucrose-shift experiment did not, however, involve intentional confounding of the number of daily trials with changes in reward magnitude, but such a procedure is common in reward-sequence experiments involving shifts (e.g., Leonard, 1969).…”