Date of this Version
Conditioned suppression was used with rats to study the effects of extending conditioned stimuli (CSs) before versus after the delivery of unconditioned stimuli. These extensions are termed B and A extensions, respectively. Within-group designs were used to compare the effects of extending CSs when 2-min parts of those CSs were separated by temporal gaps of 6 min versus a separation of no gap. The results were as follows: (a) B extensions weakened conditioning more than did A extensions, with or without gaps; (b) under some conditions, this asymmetrical effect persisted with extended training; (c) gaps between 2-min parts of a B extension had no detectable effect; and (d) under some parameter values, gaps between 2-min parts of an A extension weakened conditioning significantly. These results are better predicted by the Sometimes Opponent-Process model (SOP; A. R. Wagner, 1981) than by the Rescorla-Wagner-Frey-Sears real–time model (J. J. B. Ayres, M. Albert, & J. C. Bombace, 1987).