An experimental analysis of a coding response theory of discrimination learning

The present study has proposed four functional rules for a coding response theory of discrimination learning. Two of these rules, 1) coding responses occur in the presence of experimental stimuli and produce discriminative stimulus consequences, and 2) code items are values or a particular stimulus dimension, were assumed by the experimental conditions. The other two rules, 1) coding responses might be any artibrary response topography, and 2) coding is necessary for discrimination learning, are tested by the experiment.

Eighteen first grade and 65 college level subjects were exposed to conditional discrimination problems under one of three coding conditions, an explicated manual code, an explicated verbal code or a no explicated code condition. Subjects using either the verbal or the manual coding forms learned at approximately the same rate, supporting the rule stating that coding response topographies may be arbitrary. No first grade non-explicit subject ever learned the conditional discrimination. This finding supports the rule asserting necessity of coding. However, all college level non-explicit subjects learned, and at a rate equal to that of the two explicit coding groups. An analysis of verbal reports taken at the end of training for the non-explicit college subjects revealed that all of these subjects reported using some form of self-initiated coding. This finding adds support to the assertion that coding is a necessary part of discrimination learning.

The results of the study support the plausibility of an objectively based coding response theory of discrimination learning. Such an account readily lends itself to further research and to practical applications.