Date of Award:


Document Type:


Degree Name:

Doctor of Philosophy (PhD)



Committee Chair(s)

J. Grayson Osborne


J. Grayson Osborne


Frank Ascione


Edward Crossman


Charles Salzberg


Sebastian Striefel


This research asked whether a contextual stimulus in a visual conditional discrimination task controlled membership in classes of stimuli related hierarchically. Six experiments with nonreading preschool children posed the following question:

Does a stimulus juxtaposed with a conditional discrimination task control relations among the stimuli involved in the task? In Experiments I and II, printed instance or concept words were juxtaposed with conditional discrimination tasks involving symbols. Results for eight of nine children demonstrated neither conditional nor equivalence relations between words and symbols.

Would conditional discrimination training establish classes of visual stimuli composed of selectively nonequivalent subsets? In Experiment III, subjects from the first two experiments were taught conditional relations, then tested for stimulus class development. Printed words that could have been related transitively were not, apparently due to interference by identical letters in certain words, so no stimulus classes developed.

Would the equivalence relations sought in Experiment III develop without a history where printed words were unnecessary to conditional discrimination tasks? For Experiment IV-A, one experimentally naive child was taught the same conditional relations as Experiment III subjects. Two stimulus classes emerged, each containing two subsets that were selectively nonequivalent depending upon trial context.

Are direct or transitive stimulus relations more likely to control responding? In experiment IV-B, the subject from Experiment IV-A expressed more direct than transitive relations on modified matching trials.

Would interference by identical elements in words be precluded by training conditional relations among words directly? One child in Experiment V was taught conditional relations between concept words and instance words, and instance words and symbols. Results suggested that stimulus class development, which would have answered the question affirmatively, had begun but was incomplete.

Would providing auditory labels for some printed words preclude interference by identical elements, allowing nonidentical words to be related transitively? In Experiment VI, one child was taught auditory labels for selected printed words, followed by the same visual conditional discrimination training provided in Experiment III. Two stimulus classes developed, requiring transitive and symmetric relations among printed words.



Included in

Psychology Commons