The respondent-type training procedure and derived relational responding in adults and children
University College Cork
The training procedures employed in the investigation of stimulus equivalence typically incorporate some form of operant requirement. The purpose of this thesis was to develop a non-operant training procedure that successfully produces equivalence in human populations. In Chapter 1, a brief literature review and rationale for the thesis is offered. In Chapter 2, the effectiveness of the respondent-type training procedure was investigated with adult subjects. Nine nonsense syllables were presented to the subject in the form of six stimulus pairs on a computer screen. The first stimulus of each pair was presented on the screen for 1-s. The screen then cleared for 0.5-s ( the within-pair-delay) and the second stimulus of the pair appeared for 1-s. The screen then cleared for 3-s (the between-pair-delay) before the next stimulus pair was presented. All six stimulus pairs were presented in this fashion in a quasirandom order across 60 trials. Subjects were then presented with a standard matching-to-sample equivalence test, which tested for the emergent symmetry and equivalence relations. The majority of subjects reliably demonstrated equivalence responding after two, three of four exposures to the training and testing. The experiments reported in Chapter 3 investigated the effects of varying the within- and between-pair-delays. It was demonstrated that the respondent-type training procedure is effective using a combination of within- and between-pairdelays and is not limited to the relatively short delays used in Chapters 2 and 3. It was also demonstrated that decreasing stimulus presentations, and increasing class size, did not adversely effect performance on equivalence tests. In Chapter 4, the respondent-type training procedure was systematically compared to the matching-to-sample training procedure. On a trials for trial basis the respondent-type training procedure was shown to be more effective than the matching-to-sample training procedure. The latter procedure was shown to be as effective as the respondent-type training procedure when negative comparisons during matching-to-sample training were removed. Having identified the presence of negative comparisons as a source of disruption in the formation of equivalence classes using the matching-to-sample procedure, the experiments reported in Chapter 5 examined the effects of manipulation the functions of the negative comparisons on equivalence formation. Specifically, during one-to-many matching-to-sample training, the comparisons (e.g., Bl-B2, Cl-C2) are always presented together (comparison contiguity). When the stimulus configuration is manipulated during a matching-to-sample test, children are more likely to respond in accordance with contiguity while adults respond in accordance with the trained relations. The experiments reported in Chapter 6 demonstrated that 5 year old children reliably form equivalence using the respondent-type training procedure. In this chapter class size was increased to a four member class and the effectiveness of the procedure using a linear, one-to-many and many-to-one protocol was also investigated. In Chapter 7, five year old children were successfully taught to form fraction-decimal equivalence using the respondent-type training procedure. Subjects also showed generalisation based of physical similarity. In Chapter 8, the theoretical issues arising from this thesis are discussed.
Respondent-type training procedure , Derived relational responding , Stimulus equivalence , Operant requirement
Leader, G. 1999. The respondent-type training procedure and derived relational responding in adults and children. PhD Thesis, University College Cork.