The fate of redundant cues: Further analysis of the redundancy effect.

Peter M Jones,John M Pearce

doi:10.3758/s13420-014-0162-x

Peter M Jones, John M Pearce

Open Access

https://doi.org/10.3758/s13420-014-0162-x

Copy DOI

Journal: Learning & Behavior	Publication Date: Dec 24, 2014
Citations: 11	License type: CC BY 4.0

Affiliation: Cardiff University, University of Plymouth

Abstract

Pearce, Dopson, Haselgrove, and Esber (Journal of Experimental Psychology: Animal Behavior Processes, 38, 167–179, 2012) conducted a series of experiments with rats and pigeons in which the conditioned responding elicited by two types of redundant cue was compared. One of these redundant cues was a blocked cue X from A+ AX+ training, whereas the other was cue Y from a simple discrimination BY+ CY–. Greater conditioned responding was elicited by X than by Y; we refer to this difference as the redundancy effect. To test an explanation of this effect in terms of comparator theory (Denniston, Savastano, & Miller, 2001), a single group of rats in Experiment 1 received training of the form A+ AX+ BY+ CY–, followed by an A– Y+ discrimination. Responding to the individual cues was tested both before and after the latter discrimination. In addition to a replication of the redundancy effect during the earlier test, we observed stronger responding to B than to X, both during the earlier test and, in contradiction of the theory, after the A– Y+ discrimination. In Experiment 2, a blocking group received A+ AX+, a continuous group received AX+ BX–, and a partial group received AX± BX± training. Subsequent tests with X again demonstrated the redundancy effect, but also revealed a stronger response in the partial than in the continuous group. This pattern of results is difficult to explain with error-correction theories that assume that stimuli compete for associative strength during conditioning. We suggest, instead, that the influence of a redundant cue is determined by its relationship with the event with which it is paired, and by the attention it is paid.

Highlights

Pearce, Dopson, Haselgrove, and Esber (Journal of Experimental Psychology: Animal Behavior Processes, 38, 167–179, 2012) conducted a series of experiments with rats and pigeons in which the conditioned responding elicited by two types of redundant cue was compared
The continuous group received a simple discrimination in which AX was always followed by the unconditioned stimulus (US), and BX was never followed by the US (AX+ BX–)
According to Rescorla and Wagner (1972), for example, conditioned stimuli (CS) compete for a limited amount of “associative strength,” and this competition restricts what is learned about redundant cues

Summary

Introduction

Dopson, Haselgrove, and Esber (Journal of Experimental Psychology: Animal Behavior Processes, 38, 167–179, 2012) conducted a series of experiments with rats and pigeons in which the conditioned responding elicited by two types of redundant cue was compared. Kamin found that if X in the blocking group provided some extra information about the US (achieved by increasing the intensity of the US on AX+ trials), X gained substantial control over subsequent behavior, even after initial A+ training. In the case of blocking, initial A+ training allows A to gain substantial associative strength, restricting the quantity available on subsequent AX+ trials and leaving X with very little associative strength This account makes a clear prediction about the relative associative strengths of the redundant cues in blocking and relative-validity tasks: An AX+ BX– discrimination will enable X to gain some associative strength, whereas A+ AX+ training will lead to X having no associative strength, provided that A+ training is complete. According to this analysis, blocking should be a more effective means than the simple discrimination for keeping to a minimum the associative strength of the redundant cue

Objectives

Methods

Results