Association for Behavior Analysis International

Research on stimulus equivalence and transfer of functions opened new perspective for a behavioral account of language, cognition, and symbolic behavior. Research has shown that equivalent stimuli are substitutable for each other, sharing meaning and being semantically related. Most studies on stimulus equivalence investigate the formation of classes comprising presumably meaningless stimuli, such as abstract pictures or nonsense syllables or pseudo-words. The four presentations of this symposium will report experimental studies using meaningful stimuli to investigate issues such as transfer of functions, stability of classes, delayed emergence, and false memories. The first presentation, by Eilertsen and Arntzen, investigated transfer of functions in three classes, each comprising a facial expression of emotion (either anger, neutrality, or happiness). The degree of transfer, assessed by a Semantic Differential, was verified for three training conditions: with simultaneous matching, 0-delay matching, and 3-s delayed matching. Transfer was enhanced for both 3-s matching and simultaneous matching. The second presentation, by Silveira et al., will report an experiment in which three classes were formed with emotional expressions and meaningless stimuli, after training with either simultaneous or delayed matching. Maintenance of the classes was assessed 30 days after initial formation, and results confirm a "happy superiority effect", showing more precise responding for stimuli in the classes involving happy faces. The third study, by Augland, Lian, & Arntzen compared formation of equivalence classes comprising only abstract stimuli, and classes comprising meaningful stimuli, to investigate variables that could provide a better understanding of the phenomenon of delayed emergence of equivalence classes. The last presentation, by Aggio, Zanesco, and de Rose, investigated the role of stimulus equivalence comprising meaningful pictures and nonsense words in the production of semantic false memories, characterized by intrusions of equivalent stimuli in the recall of items in a list of nonsense words. Robust false memories were observed when memory tests were conducted seven days after formation of the classes, when stability of the classes was assured by a baseline review before the tests.

In this experiment we verified the long-term maintenance of equivalence classes established by Delayed Matching-to-Sample (DMTS) and Simultanous Matching-to-Sample (SMTS) procedures. The participants were trained to relate pictures of human faces portraying anger (A1), happiness (A2), and neutrality (A3); and abstract stimuli (B1, B2, B3, C1, C2, C3, D1, D2 and D3). Thirty days after the establishment of A1B1C1D1, A2B2C2D2 and A3B3C3D3 classes, the participants were re-exposed to equivalence probes BD and DB. We did not find a significant difference on the probability of correct responses during BD and DB maintenance tests between the participants trained with DMTS (N=15) and SMTS (N=15). However, we observed that correct responses were more likely for the BD and DB relations associated with the pictures A2 (happy faces). Our findings indicate that the long-term maintenance of equivalence classes is more sensitive to the stimulus valence rather than to participants’ prior experience with SMTS and DMTS.

The aim of the present line of experiments was to shed light on the role of trial types on delayed emergence in test for equivalence class formation. In Experiments 1 and 2 participants underwent concurrent training with meaningful pictures and abstract C-stimuli, respectively. All participants in these experiments underwent tests both with and without baseline trials interspersed in test for emergent relations. None of the participants experiencing training with abstract C-stimuli responded in accordance with stimulus equivalence (see Table 1). In the experiment where meaningful pictures were used as C-stimuli, all participants responded in accordance with stimulus equivalence. With these data, it is hard to say anything on regards to which variables influence delayed emergence. An Experiment 3 with serialized introducing of baseline trials, which has proved to have a positive outcome on stimulus equivalence formation, is therefore planned. Manipulations of baseline trials will be the same as for Experiments 1 and 2. It is finally planned an Experiment 4 where variables during training will be the same as for the first two experiments. Manipulations where transitivity and equivalence trials will not be tested for, will hopefully give us a better understanding of why we sometimes see delayed emergence of stimulus equivalence.

False memories have been studied using the Stimulus Equivalence Paradigm. The stability of equivalence classes has been suggested as an important variable for the phenomenon of false memories. In this experiment participants learned three 4-member and three 12-member equivalence classes comprising nonsense words and one familiar stimulus in each class. One week later participants retrained baseline relations from all equivalence classes and then saw a study list with nine of the 11 nonsense words from each equivalence class. After 3 minutes they were asked to recognize these stimuli from a list comprised of all stimuli from the study lists (targets), two other stimuli from the same classes (critical lures) and three stimuli from each of the other three classes (non-related stimuli). Stability of equivalence classes was measured comparing performance in sorting tests conducted before and after participants learned equivalence classes and one week later. Seven participants have already finished the experiment. Participants recognized as being in the study list 76% of the targets, 69% of critical lures and 38% of non-related. Therefore stimuli that were from the same class of the targets were more falsely recognized than stimuli from other classes. Sorting test showed stability of equivalence classes.