Research on the derived transformation of stimulus functions (ToF) typically employs single dependent measures for assessing the stimulus functions after derived relations have been established. For the first time, we examined ToF using three dependent measures both prior to and after relational training and testing. Specifically, we employed self-reports, implicit association tests, and frontal alpha asymmetry as pre versus post measures for assessing ToF. First, we trained two abstract shapes as contextual cues for happier-than and unhappier-than relations, respectively. Next, four conditional discriminations (A+/B-, B+/C-, C+/D-, and D+/E-) were trained in the presence of the happier-than cue only, where A, B, C, D, and E were blurred faces. This was followed by tests for contextually controlled transitive inference (TI) in the presence of both the happier-than and unhappier-than cues. For the participants who demonstrated TI, performance across all three measures following relational training and testing indicated that the Bhappiness^functions of the A/B stimuli were greater than those of the D/E stimuli. This constitutes the first known demonstration of emotional ToF along explicit, implicit, and neurophysiological measures concurrently.
Keywords Transformation of functions . Frontal alpha asymmetry . Transitive inference . Symbolic relationsVarious organisms can infer a relation between a wide array of stimuli given a common, mediating stimulus, and in so doing demonstrate what has been called a Btransitive inference^ (TI;Vasconcelos, 2008). To illustrate TI in a human, imagine we provide a participant with the following statements-BAdam is happier than Bob,^BBob is happier than Carry,^BCarry is happier than Dan,^and BDan is happier than Eden.^If he or she were then asked to select the Bhappier^individual from between Bob and Dan, selecting Bob would be deemed a demonstration of TI, since to provide the correct answer required making an inference based on the common node (Carry) that Bob and Dan are both related to (i.e