Abstract
Earlier research indicated that errors were more important in discrimination learning than correct responses (the Moss-Harlow effect). This conclusion was based on a variety of experimental designs each of which had basic methodological problems, usually involving novelty or order effects. A circular apparatus which allowed the simultaneous presentation of several different stimuli was used to study discrimination learning with S + and S − differentially weighted. It was found that variability in S − was more detrimental to discrimination learning than variability in S + although this finding depended on the presence of a constant stimulus. The results were interpreted as supporting the proposition that in discrimination learning rats learn more to avoid S− than to approach S+.
Get full access to this article
View all access options for this article.
