NODAL DISTANCE EFFECTS IN STIMULUS EQUIVALENCE: AN

SANCHEZ, FEDERICO JOSÉ; TABULLO, ÁNGEL; ARISMENDI, MARIANA; SEGURA, ENRIQUE; YORIO, ALBERTO

Córdoba

Congreso; IRCN First Joint Meeting of the Argentine Society for Neurosciences (SAN) and the Argentine Workshop in Neurosciences (TAN) Huerta Grande, September 4-6, 2009; 2009

Sociedad Argentina de Neurociencias

Equivalence classes paradigm has been applied to the study of abstract categories (Sidman, 1982; Zentall, 2002). An equivalence class is learned when, after training conditional discriminations between stimuli (AB, BC, CD), untrained properties of reflexivity, simmetry and transitivity emerge. It has been proposed that equivalence classes constitute an associative network, where each association between stimuli correspond to a link between two nodes. (Fields, Adams, Verhave and Newman, 1990). It has been observed that, after training AB, BC, CD associations, accuracy is higher and response times are lower for tests of derived relations separated by one node (AC, BD) than for relations separated by 2 nodes (AD, BD). This is known as “nodal distance effect”. The current experiment studied nodal distance on behavioral measures and event-related potentials. 86 subjects learned two five-member equivalence classes (A,B,C,D,E) by matching to sample. Half the subjects were trained in an ordered sequence (AB-BC-CDDE) and half in a non-ordered sequence (CD-DE-AB-BC). Derived relations of 1 (AC, CA), 2 (DA, DA) and 3 (EA, AE) nodes were tested. Percentage of correct responses was lower on tests of 3 node relations (p 0.001). The procedure was repeated on 17 subjects, testing for derived relations of 0, 1, and 3 nodes while measuring EEG activity. A positive, P300-like ERP was found between 400-600ms. The P300 was higher for the greatest nodal distance (p = 0.041), but only on subjects trained in an ordered sequence. Results suggest that relations with higher nodal distance pose greater cognitive demands, because they require activation of a larger number of elements in the associative network of the equivalence class.