Secondary reinforcement as affected by reward schedule and the testing situation.

Publisher Website
Google Scholar
Abstract
Relatively little is known about the effect of a random partial-reinforcement schedule on the strength of a secondary reinforcer. Aside from an early paper by Saltzman (8) and a recent report (5), there is virtually no other published work directed specifically to this problem. In view of the obvious importance of this variable, it would seem desirable to obtain additional relevant data. A second, concurrent purpose is to determine whether the disparity between "single-group" and "separate-groups" designs noted earlier (2) extends to this variable. Previous work (4) had indicated that amount of primary reward does not significantly affect the strength of a secondary reinforcer when separate groups of Ss are used for the high and low reward conditions. However, it was later reported (2) that if the same S is made to choose between two cues formerly associated with different amounts of primary reward (single-group design), a distinct preference develops for the stimulus previously correlated with the relatively large reward, indicating that under the latter conditions amount of reward is a significant variable. Both types of designs are utilized in the present report. Experiment I: In this experiment the separate-groups design was employed. One group of Ss received 100% reinforcement in a characteristic end box of a straight alley, while a second group was rewarded on only 50% of the trials, randomly. It has been suggested that the former schedule should result in a secondary reinforcer with greater "instrumental value" (7), or one leading to a more rapid rate of learning of a new response (6), but one, nevertheless, which succumbs to extinction more rapidly than a secondary reinforcer established under partial reinforcement. Sufficient testing trials have been included in the present experiment to produce data bearing on this question. Experiment II: In this experiment, a single-group design, each animal serves as its own control. For each S one secondary reinforcer is established under continuous reward training, while a second is established under partial reinforcement; during testing, S is required to choose between the two. Reasoning from the results of Experiment I, we might expect that for the first several testing trials (or even for the first testing day) the choices would be rather evenly divided, with a preference subsequently emerging for the secondary reinforcer developed under the partial-reinforcement schedule. (PsycINFO Database Record (c) 2006 APA, all rights reserved)

This publication has 0 references indexed in Scilit: