Physiology & Behavior, Vol. 35, pp. 897-904. Copyright ©Pergamon Press Ltd., 1985. Printed in the U.S.A. 0031-8384/85 $3.00 + .00 Reward, Performance, and the Response Strength Method in Self-Stimulating Rats: Validation and Neuroleptics AIMEE L. HAMILTON, JAMES R. STELLAR 1 AND ELIZABETH B. HART Department of Psychology and Social Relations, Harvard University, 33 Kirkland Street Cambridge, MA 02138 Received 12 December 1984 HAMILTON, A. L., J. R. STELLAR AND E. B. HART. Reward, performance, and the response strength method in self-stimulating rats: Validation and neuroleptics. PHYSIOL BEHAV 35(6)897-904, 1985.--The response strength method consists of exposing the subject to a series of variable interval schedules of reinforcement at differing densities. Response rate is plotted against obtained reinforcement rate for each schedule. The data conform to a negatively acceler- ated curve that is fit well by an analytical representation which contains two parameters. The values of these parameters are obtained from the fitted curve, and are suggested to independently reflect reinforcement and performance functions. In a first experiment, two manipulations were conducted that validated these suggestions. First, lowering the frequency of brain stimulation pulses induced a relatively selective shift in the reinforcement parameter. Second, increasing the force required to press the lever primarily altered the performance parameter. In a second experiment, the effects of neuroleptic administration on these two parameters were noted and compared to the results of the first experiment. In general, neuroleptics were seen to produce both reward and motor/performance impairments in self-stimulating rats. Self-stimulation Reward vs. performance Response strength Neuroleptics ONE of the chief problems in the study of the neural mech- anisms underlying intracranial self-stimulation (ICSS) of the brain is the difficulty of determining whether a particular manipulation has altered the rewarding effects of the stimu- lation or the capacity of the animal to respond for the brain stimulation, or both. The problem of dissociating these fac- tors affecting ICSS is particularly acute in the study of dopamine-blocking neuroleptic drugs because dopamine has been implicated in both motor and reward processes [17,28]. To address this problem, a number of techniques have been invented such as the extinction method [9], the set-reset method [21], and the more quantitative reward summation function method [7,26]. For a fuller discussion of this issue see [27]. We now introduce to brain stimulation research, a new technique which shares with the reward summation function method the advantage of yielding quantitative measures of both reward and performance capacity. Additionally, be- cause this method is derived from a long line of operant research based on Herrnstein's matching law [6,13], it has a theoretically well-grounded analytical equation that can be fit to the data. This permits two further advantages: (1) the reward and performance measures are derived from the entire data set and not just part of it as is the case with previous methods, and (2) it is possible to generate a statistic to tell if the procedure is working properly or if the experi- menter has pushed the method beyond its working limits (e.g., drug dosage is too high). ~Requests for reprints should be addressed to James R. Stellar. The response strength method, like the matching law from which it is derived, assumes that the level of perform- ance of a lever pressing operant depends upon the strength of the reinforcement for that operant relative to all other rein- forcements available to the subject [6, 13, 14]. Adapted to a single lever situation the analytic expression is: B=k(R)/R+Re, where B is the measured response rate, R is the measured reinforcement rate obtained by response rate B, k is the performance parameter, and Re is the reinforce- ment parameter. The parameter k reflects the maximum per- formance capacity of the subject under the existing test conditions. This can be seen by noting that when the rein- forcement rate R is so high that it completely dwarfs all other reinforcements available to the animal (Re), the equation reduces to B=k (because Re becomes insignificant and the R components cancel). The Re parameter indicates the rein- forcement effectiveness by measuring how much brain stimulation is necessary to get the subject to expend half of its capacity on the lever press response. Note that when the parameters R and Re are equal, the equation reduces to B=k/2 or half maximal response level. Thus, the reinforce- ment required to sustain half-maximal responding (Re) is a marker for the effectiveness of brain stimulation in generat- ing reinforcement. An increase in Re would indicate a de- crease in the reinforcement effectiveness of the brain stimu- lation. In the domain of natural reinforcements (e.g., food), ex- periments have been conducted with this method to test the 897