Consequences of response-contingent change in unconditioned stimulus intensity upon the rabbit (Oryctolagus cuniculus) nictitating membrane response.

In a between-groups design, an instrumental contingency was superimposed on classical conditioning of the rabbit nictitating membrane response. Subjects were exposed to a tonal conditioned stimulus (CS) and, if no conditioned response (CR) occurred, a 5.0-ma. paraorbital electric shock unconditioned stimulus (US). USs of 5.0, 3.3, 1.7, and 0.0 ma., respectively, were contingent on occurrence of a CR in the CS-US interval for subjects in the 4 groups of the experiment. The group exposed to contingent US omission differed from the other 3 groups in percentage CR, onset latency, amplitude, and 2 indices of CR form, and those 3 groups generally did not differ significantly among themselves on these dependent variables. Results were interpreted to be contrary to "law of effect" formulations of classical conditioning.