The paradoxical enhancement of classically conditioned physiological responses in human subjects

Abstract

This study investigated the phenomenon that, under some conditions, presentation of a Conditioned Stimulus (CS) alone produces an "enhancement" of the Conditioned Response (CR) instead of the well-known extinction effect. Three groups of five male undergraduates each received CS (slide presentations) - UCS (0.5 second burst of white noise) pairings on a 76 per cent irregular reinforcement schedule. Group treatments differed by the number of CS-alone presentations given immediately after conditioning (either 4, 12 or 30). CS-alone and UCS-alone control groups were also used. At two extinction sessions, a week apart, the magnitude (a change of at least 100 ohms between 0.9 and 5.0 seconds after CS onset) and latency of the Galvanic Skin Response (GSR) and change in Finger Pulse (FP) rate were recorded. The primary results showed: (a) a significantly higher GSR magnitude for the conditioning groups as compared to the control groups (F(4, 20) = 3.87, p < .025); (b) the group with the greatest number of CS-alone presentations after conditioning was significantly different from the other experimental groups of GSR magnitude (F(2, 20) = 7.11, p < .05) and FP rate (p < .05; Duncan's Multiple Range Test); (c) this same group did not extinguish after 60 CS-alone presentations while the other experimental groups extinguished after approximately 20 CS-alone presentations; and (d) a significant Trial Block x Group interaction for the latency of the GSR was noted (F(20, 100) = 1.68, p < .05). Some implications of these findings and parameters for future research are discussed.