Skip to main navigation Skip to search Skip to main content

Human choice and the matching law: the role of verbal behaviour

    Student thesis: Doctor of Philosophy

    Abstract

    Differences between the performances of animals and humans on basic schedules of reinforcement are now well-documented (Lowe , 1979, 1983; Weiner, 1970b, 1983). Moreover, considerable evidence has emerged to suggest that rule-governed behaviour may be strongly implicated in this disparity. A number of studies, however, have reported that, like animals, human performance on multiple variable interval (VI), and concurrent VI, schedules is
    well-described by Herrnstein's single rate equation and the generalised matching equation, respectively.
    In an attempt to resolve this discrepancy the determinants of human performance on multiple VI, multiple and simple concurrent VI, schedules were evaluated and compared with animal data. Variables manipulated were: (i) changeover-delay (COD), (ii) schedule cues and (iii) schedule instructions. In the
    concurrent schedule experiments only schedule cues and schedule instructions enhanced the likelihood of matching-like performances. Overall, however, only 17 out of 40 subjects' data was described by the generalised matching equation with parameter estimates commensurate with those reported in animal studies. Moreover, the data of only 13 subjects could be described by both of Herrnstein's single rate equations. In the multiple VI experiments, 3 out of 10 subjects' data could not be described by Herrnstein's single rate equation. Pigeon performance, however, on the same multiple VI and multiple
    concurrent VI schedules was well-described by Hermstein's single rate equation and the generalised matching equation, respectively.
    Human subjects' retrospectively reported performance rules were consistent with their patterns of behaviour allocation. Only those subjects who enunciated matching-like response rules performed in a manner which conformed to the generalised matching equation and compared well with animal data.
    Animal-human differences in schedule performance may therefore be extended to include multiple, and concurrent, VI schedules and evidence from subjects' verbal reports is consistent with the view that these differences depend upon the human capacity for rule-governed behaviour.
    Date of Award1986
    Original languageEnglish
    Awarding Institution
    • University College of North Wales, Bangor
    SupervisorFergus Lowe (Supervisor)

    Cite this

    '