MURAL - Maynooth University Research Archive Library

    Developing the Function Acquisition Speed Test: Using a Functional Research Approach to Build a Novel Implicit Test

    O'Reilly, Anthony (2012) Developing the Function Acquisition Speed Test: Using a Functional Research Approach to Build a Novel Implicit Test. PhD thesis, National University of Ireland Maynooth.

    Download (1MB) | Preview

    Share your research

    Twitter Facebook LinkedIn GooglePlus Email more...

    Add this article to your Mendeley library


    The eleven studies reported in this thesis outline the development of a novel implicit test for assessing verbal and social histories. This measure was named the “Function Acquisition Speed Test” (FAST). The current research utilizes a functional research approach drawing upon the seminal research by Watt, Keenan, Barnes and Cairns (1991) and upon more recent research by Gavin, Roche and Ruiz (2008) to inform the bottom-up development of the FAST. Chapter 1 presents a review the behaviour analytic literature concerned with the phenomena of stimulus equivalence and derived relational responding, and reviews the links between these concepts and complex human behaviour such as attitudes. The seminal study by Watt et al. (1991) is also described. In that study, the researchers attempted to train subjects to form two equivalence classes which were incongruent with Northern Irish subjects’ socioverbal history (i.e. containing Catholic names and Protestant symbols) using a stimulus equivalence paradigm. Watt et al. concluded that pre-existing social histories can interfere with the acquisition of novel stimulus relations. The relevance of this work as an underpinning principle of the FAST is described. Chapter 1 also briefly reviews the social cognitive approach to implicit testing and discusses the value of a functional approach. Also in Chapter 1, the initial structure of the FAST paradigm is outlined. The key FAST test blocks present subjects with one stimulus per trial, and require subjects to learn to emit one of two responses (i.e. press right or press left) based on corrective feedback. In the consistent block, the same response is required for the two stimuli suspected to be related, while the other response is required for two novel, unrelated stimuli. In the inconsistent block, the two stimuli of interest require two different responses. The reinforcement contingencies of the inconsistent act as a behavioural disrupter, and the difference in learning rates between the two blocks is taken to be indicative of the strength of the relation between the test stimuli. Chapter 2 reports on five experiments that tested the underlying premise of the FAST with regard to experimentally trained stimulus relations. In each of the experiments in this chapter, subjects completed training procedures to establish a stimulus relation which would later be probed for using the FAST methodology. In Experiment 1, subjects completed matching to sample training which established a simple A-B relation between two three-letter nonsense syllables. The FAST was capable of detecting the trained relations in a majority of subjects and on a group level, thus establishing a basic proof of concept for the methodology. Experiment 2 sought to expand on this by establishing aversive or erotic stimulus functions for a pair of nonsense syllables. The FAST procedure was then used to test for a relation between a nonsense syllable and related (but novel) romantic images. The FAST was also run multiple times for each subject. The results revealed a great deal of instability over time, with the expected effect only emerging on the second of three FAST runs. In Experiment 3, subjects were trained using a Matching-to-Sample procedure to create two three-member equivalence classes. The derived relations (A1-C1) were then tested for using the FAST. Given the importance of derived relational responding to a behavioural account of attitudes, this was a vital test for the fledgling format. The FAST was successful in detecting the derived relations, and this effect emerged on all three runs of the FAST. Experiment 4 sought to further build on this success by utilising an extremely robust and extended training procedure to maximise the probability of the emergence of stable equivalence relations. The training procedure contained nine stages of increasing difficulty and fading of consequences. A repeated single subjects design was used, and 4 out of five subjects showed strong positive FAST effects when tested. Experiment 5 investigated the question of whether the FAST was capable of detecting the influence of derived relations which had not yet been tested for by a normal equivalence testing procedure. All subjects completed an equivalence training procedure, but half of the experimental subjects completed a FAST prior to completing an equivalence test, while the other half completed a FAST after completing a normal equivalence test. The FAST was unable to detect the derived relations in the former experimental group, even in subjects who latter passed the equivalence test. The implications of this result for a Relational Frame Theory account of implicit attitudes are discussed. In Chapter 3, the effect of a number of basic procedural modifications to the FAST presentation was considered. In each experiment, subjects completed 3 successive runs of a FAST after having been trained in a simple A-B stimulus relation. Experiment 6 used a between subjects design to measure the effect of varied response windows on FAST performance (1000ms, 2000ms, 3000ms or 4000ms). Experiment 7 investigated the effect of including a variable number of practice blocks prior to the first baseline block (3, 5, or 7 practice blocks). Experiment 8 varied the fluency criterion for completing a FAST block. (7, 8, or 9 correct in a sequence of ten responses.). The results of these experiments were difficult to interpret due to instability in the predicted FAST effect across the three runs of the FAST. Possible reasons for this instability are discussed, and improvements to the FAST are suggested. The results of these experiments resulted in a shortening of the FAST response window to 2000ms, the inclusion of a practice block, and the inclusion of a “counter” which displayed the number of correct responses in a row that the subject had emitted. Chapter 4 tested the FAST with “real world” stimuli as the test stimuli. In each experiment in Chapter 4, subjects complete FAST procedures aimed at detecting natural histories of verbal behaviour. In each experiment, a different relation is targeted. In experiment 9, the FAST probes for a relation between the words “spider” and “disgust”. Experiment 10 targets the relations “immigrant” and “cheat”. Experiment 11 departs from using words as stimuli, using images of teenaged females to probe for relations between sexual images and images of teenaged females of different ages. The results of these experiments demonstrate both the limitations and strengths of the FAST procedure in different contexts. The FAST was shown to be most capable of detecting culturally ubiquitous verbal relations (spiders are disgusting); particularly in cases where the two stimuli are strongly in opposition (i.e. young girls are sexual). Experiment 11 also demonstrated the potential of the FAST to be deployed in serial to examine the “shape” of an equivalence class. Subjects in experiment 11 showed a strong negative FAST effect when female stimuli were pre-teenaged, but this effect quickly vanished as the ages of the pictured females increased into the teenage years. Chapter 5 provides an overview of the entire research project. The experimental findings from each of the previous chapters are discussed along with the implications of these findings. The early development of this novel test format is charted, and remaining challenges that must be confronted in future research are outlined. Alternative experimental preparations emerging from this research are also considered and explored, and possible applications in both basic research and applied setting are discussed.

    Item Type: Thesis (PhD)
    Keywords: Function Acquisition Speed Test; Functional Research Approach; Build a Novel Implicit Test;
    Academic Unit: Faculty of Science and Engineering > Psychology
    Item ID: 5396
    Depositing User: IR eTheses
    Date Deposited: 16 Sep 2014 10:35

      Repository Staff Only(login required)

      View Item Item control page


      Downloads per month over past year