Cue-response Separation and Element Proximity in the Feature Discrimination Paradigm

75 leaves. Advisor: William D. KlipecThe problem. To investigate the function of the proximity of elements in a feature discrimination paradigm on the feature-positive effect, remote responding and stimulus identification. Procedure. Sixty preschoolers between the ages of three and five years were trained to discriminate between two simultaneously presented displays containing either four common features, or three common and one distinctive feature. One-third of the subjects were reinforced for touching any feature on the display with the distinctive feature (FP-l group), one-third for touching a common feature on the distinctive feature display (FP-2 group), and one-third for touching any feature on the common feature display (FN group). Half of each group of subjects were trained with the features compacted in the centers of the displays and half with the features distributed in the outer corners of the displays. Findings. A clear feature-positive effect occurred only when the elements were distributed. Compacting the elements resulted in fewer errors but did not eliminate a significant difference between the FP-2 and FN group. Generalization tests given during extinction indicated that subjects responded on the basis of features, pattern, or both independent of the condition to which they were assigned. Conclusions. Responses to features were influenced by reinforcement probability but responses did not have to be directed to a feature for it to affect responding. The hypothesis that cue-response separation contributes to feature-negative difficulty was not supported. Recommendations. Future research in this area should attempt to clarify the parameters of the effect, remote responding strategies, and factors that affect responding