The Functional-Analytic Development of a Test for Behavioural History using the Concept of Derived Stimulus Relations
Gavin, Amanda (2008) The Functional-Analytic Development of a Test for Behavioural History using the Concept of Derived Stimulus Relations. PhD thesis, National University of Ireland Maynooth.
Ten studies reported in this thesis outline the development of a behavioural test for assessing the role of verbal behaviour in identifying personal and social histories. The testing format was developed by employing the Stimulus Equivalence paradigm (Sidman, 1971) and extending upon the previous work of Watt, Keenan, Barnes and Cairns (1991). Chapter 1 presents a review of the behavioural literature concerned with stimulus equivalence and derived stimulus relations. Most importantly, a seminal study by Watt et al. (1991) is outlined. In that study, a simple stimulus equivalence paradigm was used to take advantage of the fact that people in Northern Ireland often respond to each other’s names as indicative of religious background. Specifically, the researchers attempted to teach subjects the necessary baseline conditional discriminations to form two equivalence classes that were incongruent with the subject’s social history. Watt et al. (1991) concluded subjects’ personal and social histories interfere with their ability to derive specific equivalence relations in the laboratory. The relevance of this paradigm to all of the experimental work reported in this thesis is outlined. Chapter 2 reports on two experiments (Experiments 1 & 2) that tested the applicability of the Watt et al. paradigm as a tool for assessing personal and social histories as discussed in Chapter 1. In Experiment 1, a novel Yes/No procedure was employed in a controlled experimental laboratory preparation to create and test for social histories. This novel YES/NO procedure required subjects to respond to two stimuli in the presence of the question “Do they go together” by clicking on either a Yes or No button on a computer screen. Experiment 2 involved an experiment that expanded upon this technique by applying this novel Yes/No adaptation of the Watt et al. paradigm in an effort to assess the social categorisation of real-world terms by men and women from the general population. Both studies demonstrated the applicability of the adapted Watt et al. paradigm in assessing both laboratory-controlled and real-world social and personal histories. In Chapter 3, Experiment 3 explored the possibility of assessing differences in social history when using a novel instruction-based relational test. The test measure was capable of identifying subjects’ laboratory created histories on the basis of response accuracy differentials across the test blocks in the absence of the equivalence training used in the Watt et al. Paradigm. That is, in place of equivalence training, subjects were presented with onscreen instructions informing them which stimuli to “put together” in the relevant phase of the test. Subjects were then presented with two test blocks, accompanied by different matching instructions. These blocks consisted of matching tasks involving the presentation of word pairs and in which responses to the Yes or No buttons were required. One set of rules was congruent with the subjects’ personal/social history and the other set was incongruent with the subjects’ personal/social history. This greatly modified Watt et al. procedure did not require equivalence training, but nevertheless successfully identified subjects’ social and personal histories. In Chapter 4, the relational test procedure (described in Chapter 3) was modified slightly and applied in a real-world setting to examine and identify the use of socially sensitive verbal relations within a series of different populations. This was explored across 2 experiments (Experiments 4 & 5). Experiment 4 demonstrated the utility of the current testing procedure in detecting cultural differences across populations with regard to the historical categorisation of socially sensitive stimuli relevant to the issue of homosexuality and homophobia. The test format was modified in that subjects no longer responded using the onscreen YES and NO keys, but by pressing either a Blue or Red button onscreen in place of YES and NO, respectively. In this experiment, homosexual males from the USA and Ireland completed the modified relational test procedure consisting of two test blocks. Each block consisted of identical matching tasks but was accompanied by a distinct set of instructions. One set of instructions was congruent with the subjects’ history (i.e., gay goes with good) while the other was incongruent (i.e., gay goes with bad). The test was successful in identifying the cultural background of subjects taking the test. A slight modification of the foregoing procedure in Experiment 5 gave rise to similar results with regard to female subjects’ categorisation of terms relating to children and sex. That is, when the relational test procedure was presented as before, but with subjects responding using keys on the keyboard, female subjects' fluency in associating child and sex terms was lower than their fluency in relating adult and sexual terms. That is, when instructed matching was congruent with the female subjects’ personal and social histories (i.e. child goes with nonsexual) response accuracies were greater than when matching instructions were incongruent (i.e. child goes with sexual). In Chapter 5, the experimental focus moved towards the use of a single stimulus onscreen rather than stimulus pairs. This represented a radical departure from the procedure used in Chapters 2-4. Specifically, subjects were no longer required to explicitly match the stimuli onscreen in relation to each other. Instead, Experiment 6 sought to assess the rate of acquisition of common response functions to words considered compatible for a normal population compared to words considered incompatible for a normal population. The experiment identified gender differences in the rate of acquisition of common stimulus functions for members of distinct (e.g., child and sexual) and common (e.g., adult and sexual) verbal relations. In Chapter 6, a laboratory analogue of the new single stimulus test procedure was developed. This was developed in tandem with a behavioural analysis of the Implicit Association Test (IAT; Greenwald, McGhee, & Schwartz, 1998) due to the topographical similarity of both measures. Specifically, Chapter 6 examined the IAT test format in terms of behavioural processes whilst also providing a laboratory analysis of the current test procedure using arbitrary laboratory created stimuli across a series of experiments (Experiments 7, 8, 9 & 10). Chapter 6 showed that a laboratory history of respondent conditioning and derived relational responding is sufficient in generating an IAT effect. In addition, this effect is malleable depending on the type of stimulus equivalence testing employed (i.e., symmetry and transitivity combined, transitivity alone, or no test). Chapter 7 provides a summary of the entire research program presented in the thesis, and reviews the development of a functionally-understood model of the IAT. Some important empirical and conceptual issues that arise from the various findings are also outlined. Finally, the relationship of the current research to work outside the field of behaviour analysis is considered.
Repository Staff Only: item control page