PRELIMINARY FINDINGS ON THE EFFECTS OF SELF-REFERRING AND EVALUATIVE STIMULI ON STIMULUS EQUIVALENCE CLASS FORMATION RHONDA M. MERWIN and KELLY G. WILSON
University of Mississippi
Thirty-two subjects completed 2 stimulus equivalence tasks using a matching-to-sample paradigm. One task involved direct reinforcement of conditional discriminationsdesigned to produce derived relations between self-referring stimuli (e.g. , me, myself, I) and positive evaluation words (e.g., whole, desirable, perfect). The other task was designed to produce derived relations between selfreferring stimuli and negative evaluation wordls (e.g., unworthy, flawed, inadequate). Performance on each task was recorded via response latency and percent correct. Prior tocompletion of the equivalence tasks, subjects completed 2 self-report measures: the Outcome Ouestionnaire-45 (00-45) and the Rosenberg SelfEsteem Scale (RSE) . Subjects were divided into groups based on their 00-45 score (high or low distress) and FISE score (high or low self-esteem). Significant differences in percent correct were found between both the 00-45 groups and the RSE groups. Subjectswho reported high distress and a negative sense of self made significantly more errors on the tests for equivalence for the task that required matching self-referential stimuli with positive evaluation words .
Stimulus Equivalence Murray Sidman conducted the first systElmatic investigation of stimulus equivalence in 1971. In this well-cited study, an individual with a learning disability wastaught to identify pictures of objects given the spoken words. He was then taught to identify the written words given the spoken words. Without any additional training, the individual matched the pictures to the written words and the written words to the pictures. That is, he derived relations among the stimuli in the absence of direct reinforcement for the specific choices, responding to thestimuli as if they were equivalent. Sidman referred to this behavioral phenomenon as
Correspondence may be directed to Rhonda M. Merwin , University of Mississippi , Department of Psychology, 307 Peabody Building , University, MS 38677. (E-mail : rmblevin @olemiss.edu) .
MERWIN AND WILSON
stimulus equivalence and identified response features that indicated that a stimulus equivalenceclass had been formed (i.e. , reflexivity, symmetry, and transitivity) (Sidman, 1990). Subsequent studies have supported Sidman's observations of stimulus equivalence. Specifically, studies have repeatedly demonstrated that when an individual is taught a series of related conditional discriminations (e.g ., choose A1 in the presence of B1, choose A1 in the presence of C1), he/she often derivesadditional relations that were not explicitly taught (e.g. , choosing B1 in the presence of C1 and C1 in the presence of B1). Research has demonstrated that children as young as 17 months engage in simple relating and by 23 months they demonstrate all the necessary response features of stimulus equivalence (Upkens, Hayes, &Hayes, 1993). Investigations of stimulus equivalence have focused primarily onidentifying the conditions that reliably produce equivalence (Sidman, 1990; Sidman & Tailby, 1982) and examining central features of stimulus equivalence classes. This research has demonstrated resurgence of previously established equivalence relations (Wilson & Hayes, 1996), transfer of a variety of psychological functions among class members (Dougher, Augustson , Markham, Greenway, & Wulfert, 1994;Dymond & Barnes, 1994; Roche & Barnes, 1997), differences in acquisition between language-able and language disabled populations (Devany, Hayes, & Nelson, 1986), that equivalence classes can include stimuli of various modalities (DeGrandpre, Bickel, & Higgins, 1992), and that classes can be brought under contextual control (Kohlenberg , Hayes, & Hayes, 1991 ; Steele & Hayes, 1991). Among recent...