"Shaping" or "method of successive approximation" is a procedure which may be applied to increase the frequency of a response which has a low operant level, or it may also be used to bring about responses which have not been previously emitted. In "shaping," the experimenter initially reinforces a response which is within the behavorial repertoire of the subject. Then, the experimenter reinforces only responses which approximate the behavior which is desired. The final behavior is then directly reinforced.

A procedure in which the S- was prevented from being responded to, by electro-magnets, was used to establish color discriminations. The procedure was modified in Situation 1, to include the prevention of responses to the S+ if the S- was responded to first. The original procedure and modified procedure were used in Situation 1, with only the modified procedure being used in Situations 2 and 3. The procedure of reinforcing responses to the S+ and extinguishing responses to the S-, through nonreinforcement, was used in Situation 4. Data recorded consisted of the number of trials, the number of reinforcements, and which stimulus was first responded to. Criteria for the acquisition of a discrimination was 100 first responses to the S+. Results indicated that the modified procedure was much more effective in establishing the discriminations, than the original procedure or the procedure of reinforcing responses to the S+ and extinguishing responses to the S-. The modified procedure enhanced the establishment of stimulus control, reduced the number of errors and eliminated stereotyped responses.

The investigation sought to examine resistance to extinction (Rn) as a function of previous experience with downward shifts in reward magnitude. It was suggested that previous research conducted within the framework of the Spence-Amsel frustration hypothesis and the sequential hypothesis failed to administer sufficient postshift trials to adequately establish the relationships that may exist. Under one condition, four groups of rats received twenty extinction trials following forty postshift trials. Under another condition, four groups were extinguished following eighty postshift trials. An inverse magnitude of reward effect occurred in the preshift phases, however, which prevented an adequate analysis of either the shift or the Rn data, This unexpected effect was discussed within the framework of Black's incentive-motivation interpretation of reinforcement.