Within-session rates of responding when reinforcer magnitude is changed within the session

In the present experiment, the authors investigated the idea that within-session changes in operant response rates occur because subjects sensitize and then habituate to the reinforcer. If that is true, then altering an aspect of the reinforcer within the session should alter the observed within-session responding. The authors tested that idea by having rats press a lever for 2 food-pellet reinforcers delivered by a variable-interval 120-s schedule during 60-min baseline sessions. In treatment conditions, the magnitude of the reinforcer was halved (1 pellet) or doubled (4 pellets) 10, 20, 30, 40, or 50 min into the session. That magnitude of reinforcement then remained in effect for the rest of the session. Altering reinforcer magnitude altered the rates of responding within the session in a fashion consistent with the habituation explanation, that is, response rates increased, relative to baseline, when the magnitude of reinforcement was increased. They decreased when the magnitude was decreased. Those results were seemingly inconsistent with the competing idea that within-session decreases in responding rates are produced by satiation.     

Within-session changes in responding during several simple schedules

Pigeons' key pecking was reinforced by food delivered by several fixed-interval, variable-ratio, and differential-reinforcement-of-low-rate schedules. Rate of responding, number of responses per reinforcer, length of postreinforcement pause, running response rate, and the time required to collect an available reinforcer changed systematically within sessions when the schedules provided high rates of reinforcement, but usually not when they provided low rates. These results suggest that the factors that produce within-session changes in responding are generally similar for different schedules of reinforcement. However, a separate factor may also contribute during variable-ratio schedules. The results question explanations for within-session changes that are related solely to the passage of time, to responding, and to one interpretation of attention. They support the idea that one or more factors related to reinforcement play a role.     

Within-session changes in key and lever pressing for water during several multiple variable-interval schedules

Rats pressed keys or levers for water reinforcers delivered by several multiple variable-interval schedules. The programmed rate of reinforcement varied from 15 to 240 reinforcers per hour in different conditions. Responding usually increased and then decreased within experimental sessions. As for food reinforcers, the within-session changes in both lever and key pressing were smaller, peaked later, and were more symmetrical around the middle of the session for lower than for higher rates of reinforcement. When schedules provided high rates of reinforcement, some quantitative differences appeared in the within-session changes for lever and key pressing and for food and water. These results imply that basically similar factors produce within-session changes in responding for lever and key pressing and for food and water. The nature of the reinforcer and the choice of response can also influence the quantitative properties of within-session changes at high rates of reinforcement. Finally, the results show that the application of Herrnstein's (1970) equation to rates of responding averaged over the session requires careful consideration.     

Within-session changes in responding during concurrent schedules with different reinforcers in the components.

Rats and pigeons responded on several concurrent schedules that provided different reinforcers in the two components (food and water for rats, Experiment 1; wheat and mixed grain for pigeons, Experiment 2). The rate of responding and the time spent responding on each component usually changed within the session. The within-session changes in response rates and time spent responding usually followed different patterns for the two components of a concurrent schedule. For most subjects, the bias and sensitivity to reinforcement parameters of the generalized matching law, as well as the percentage of the variance accounted for, decreased within the session. Negative sensitivity parameters were sometimes found late in the session for the concurrent food-water schedules. These results imply that within-session changes in responding could cause problems for assessing the validity of quantitative theories of concurrent-schedule responding when the components provide different reinforcers. They question changes in a general motivational state, such as arousal, as a complete explanation for within-session changes in responding. The results are compatible with satiation for, or sensitization-habituation to, the reinforcers as explanations.     

Within-session Changes In Responding During Concurrent Variable-interval Schedules

Five rats and 4 pigeons responded for food delivered by several concurrent variable-interval schedules. The sum of the rates of reinforcement programmed for the two components varied from 15 to 480 reinforcers per hour in different conditions. Rates of responding usually changed within the experimental session in a similar manner for the two components of each concurrent schedule. The within-session changes were similar to previously reported changes during simple schedules that provided rates of reinforcement equal to the sum of all reinforcers obtained from the concurrent schedules. The number of changeovers also changed within sessions in a manner similar to the changes in instrumental responding. These results suggest that changeovers are governed by the same variables that govern instrumental responding. They also suggest that the within-session change in responding during each component of a concurrent schedule is determined by approximately the sum of the reinforcers obtained from both components when both components provide the same type of reinforcer.     

Sensitization and habituation regulate reinforcer effectiveness

We argue that sensitization and habituation occur to the sensory properties of reinforcers when those reinforcers are presented repeatedly or for a prolonged time. Sensitization increases, and habituation decreases, the ability of a reinforcer to control behavior. Supporting this argument, the rate of operant responding changes systematically within experimental sessions even when the programmed rate of reinforcement is held constant across the session. These within-session changes in operant responding are produced by repeated delivery of the reinforcer, and their empirical characteristics correspond to the characteristics of behavior undergoing sensitization and habituation. Two characteristics of habituation (dishabituation, stimulus specificity) are particularly useful in separating habituation from alternative explanations. Arguing that habituation occurs to reinforcers expands the domain of habituation. The argument implies that habituation occurs to biologically important, not just to neutral, stimuli. The argument also implies that habituation may be observed in “voluntary” (operant), not just in reflexive, behavior. Expanding the domain of habituation has important implications for understanding operant and classical conditioning. Habituation may also contribute to the regulation of motivated behaviors. Habituation provides a more accurate and a less cumbersome explanation for motivated behaviors than homeostasis. Habituation also has some surprising, and easily testable, implications for the control of motivated behaviors.     

Altering Reinforcer Variety or Intensity Changes the Within-Session Decrease in Responding

Operant responding often changes systematically within experimental sessions. McSweeney, Hinson, and Cannon (1996) argued that sensitization and habituation produce within-session changes in responding. The present study tested two predictions of the sensitization–habituation explanation. In two experiments, rats pressed a lever for reinforcers delivered by a multiple variable interval 15-s variable interval 15-s schedule. In Experiment 1, the variety of reinforcers delivered during the session was manipulated by varying the percentage of programmed reinforcers replaced with qualitatively different reinforcers from 0 to 75%, in five different conditions. In Experiment 2, the intensity of the reinforcer was manipulated by varying the concentration of sucrose in the sucrose and water solution used as the reinforcer from 0 to 30%, in five different conditions. Increasing the variety or the intensity of the reinforcers slowed the within-session decrease in responding. The results are consistent with the predictions of a sensitization–habituation explanation of within-session changes in responding.     

Resistance to change and the law of effect

Three experiments using multiple schedules of reinforcement explored the implications of resistance-to-change findings for the response-reinforcer relation described by the law of effect, using both steady-state responding and responding recorded in the first few sessions of conditions. In Experiment 1, when response-independent reinforcement was increased during a third component, response rate in Components 1 and 2 decreased. This response-rate reduction was proportionately greater in a component in which reinforcer magnitude was small (2-s access to wheat) than in the component in which it was large (6-s access to wheat). However, when reinforcer rates in the two components were varied together in Experiments 2 and 3, response-rate change was the same regardless of the magnitude of reinforcers used in the two components, so that sensitivity of response rates to reinforcer rates (Experiment 2) and of response-rate ratios to reinforcer-rate ratios (Experiment 3) was unaffected by the magnitude of the reinforcers. Therefore, the principles determining resistance to change, described by behavioral momentum theory, seem not to apply when the source of behavior change is the variation of reinforcement contingencies that maintain the behavior. The use of extinction as a manipulation to study resistance to change is questioned.     

Reinforcer quality matters: A test of the Mathematical Principles of Reinforcement with domestic hens (Gallus gallus domesticus)

This study evaluated the ability of Killeen's (1994) Mathematical Principles of Reinforcement to account for the effects of changes in reinforcer quality on hens' rates of responding on fixed‐ratio schedules. Hens were trained to peck a key on a fixed‐ratio schedule of reinforcement and then experienced an ascending series of ratio values in two separate conditions. In different conditions, the food reinforcer was either wheat or puffed wheat. Response rates initially increased with increases in ratio requirement before eventually decreasing at larger ratios. Quantitative fits of the model accounted for the data well. The fits revealed that different foods were systematically associated with changes in the specific activation parameter, a, and these were consistent with previous reports of preference for those food items.     

Rapid acquisition of preference in concurrent chains when alternatives differ on multiple dimensions of reinforcement

Pigeons responded in a concurrent-chains procedure in which terminal-link reinforcer variables were changed unpredictably across sessions. In Experiment 1, the terminal-link schedules were fixed-interval (FI) 8 s and FI 16 s, and the reinforcer magnitudes were 2 s and 4 s. In Experiment 2 the probability of reinforcement (100% or 50%) was varied with immediacy and magnitude. Multiple-regression analyses showed that pigeons' initial-link response allocation was determined by current-session reinforcer variables, similar to previous studies which have varied only immediacy (Grace, Bragason, & McLean, 2003). Sensitivity coefficients were positive and statistically significant for all reinforcer variables in both experiments. Analyses of responding within individual sessions showed that final levels of preference for dominated sessions, in which all reinforcer variables favored the same terminal link, were more extreme than for tradeoff sessions in which at least one reinforcer variable favored each alternative. This result implies that response allocation was determined by multiple reinforcer variables within individual sessions, consistent with the concatenated matching law. However, in Experiment 2, there was a nonlinear (sigmoidal) relationship between response allocation and relative value, which suggests the possibility that reinforcer variables may interact during acquisition, contrary to the matching law.     

Effects of acute and chronic flunitrazepam on delay discounting in pigeons

Delay to delivery of a reinforcer can decrease responding for that reinforcer and increase responding for smaller reinforcers that are available concurrently and delivered without delay; acute administration of drugs can alter responding for large, delayed reinforcers, although the impact of chronic treatment on delay discounting is not well understood. In this experiment, the effects of repeated administration of the benzodiazepine flunitrazepam were studied in 6 pigeons responding on one key to receive food that was delivered immediately and on a second key to receive a larger amount of food that was delivered following delays which increased across a single session. Pigeons responded predominantly for the large reinforcer when there were no delays and when delays were short; however, as delays increased, responding for the large reinforcer decreased. Acutely, flunitrazepam (0.32, 1.0 and 3.2 mg/kg) dose-dependently increased responding for the large reinforcer, shifting the discounting curve rightward and upward. Repeated administration of flunitrazepam (0.32, 1.0 and 3.2 mg/kg, each for six sessions, separated by one session during which vehicle was administered) did not markedly alter its effects on responding for the large reinforcer, indicating that the development of tolerance to this effect of flunitrazepam is modest under these conditions.     

Resistance To Change As A Function Of Stimulus-reinforcer And Location-reinforcer Contingencies

Pigeons responded on two keys in each component of a multiple concurrent schedule. In one series of conditions the distribution of reinforcers between keys within one component was varied so as to produce changes in ratios of reinforcer totals for key locations when summed across components. In a second series, reinforcer allocation between components was varied so as to produce changes in ratios of reinforcer totals for components, summed across key locations. In each condition, resistance to change was assessed by presenting response-independent reinforcers during intercomponent blackouts and (for the first series) by extinction of responding on both keys in both components. Resistance to change for response totals within a component was always greater for the component with the larger total reinforcer rate. However, resistance to change for response totals at a key location was not a positive function of total reinforcement for pecking that key; indeed, relative resistance to extinction for the two locations showed a weak negative relation to ratios of reinforcer totals for key location. These results confirm the determination of resistance to change by stimulus—reinforcer contingencies.     

Effects of reinforcer consumption and magnitude on response rates during noncontingent reinforcement

Results of previous research on the effects of noncontingent reinforcement (NCR) have been inconsistent when magnitude of reinforcement was manipulated. We attempted to clarify the influence of NCR magnitude by including additional controls. In Study 1, we examined the effects of reinforcer consumption time by comparing the same magnitude of NCR when session time was and was not corrected to account for reinforcer consumption. Lower response rates were observed when session time was not corrected, indicating that reinforcer consumption can suppress response rates. In Study 2, we first selected varying reinforcer magnitudes (small, medium, and large) on the basis of corrected response rates observed during a contingent reinforcement condition and then compared the effects of these magnitudes during NCR. One participant exhibited lower response rates when large-magnitude reinforcers were delivered; the other ceased responding altogether even when small-magnitude reinforcers were delivered. We also compared the effects of the same NCR magnitude (medium) during 10-min and 30-min sessions. Lower response rates were observed during 30-min sessions, indicating that the number of reinforcers consumed across a session can have the same effect as the number consumed per reinforcer delivery. These findings indicate that, even when response rate is corrected to account for reinforcer consumption, larger magnitudes of NCR (defined on either a per-delivery or per-session basis) result in lower response rates than do smaller magnitudes.     

Within-Session Patterns of Pigeons'' General Activity

The study investigates the idea that within-session changes in responding are produced by arousal and satiation. General activity, measured by the displacement of floor panels, was used as an index of these variables. In Experiment 1, pigeons pecked a key on a simple variable-interval schedule and general activity was also recorded. In Experiment 2, pigeons received response-independent reinforcers. In Experiment 3, the pigeons' general activity was reinforced on a variable-interval schedule. Although significant within-session changes in operant key pecking were observed in Experiment 1, within-session changes in general activity were seldom observed in either Experiment 1 or Experiment 2. Significant within-session changes in general activity were observed in Experiment 3, when activity was the operant. The present results can potentially be explained by arousal and satiation. However, the predictive power of this explanation is severely limited. Regardless of whether one accepts arousal and satiation as the theoretical explanation for within-session changes in responding, the present results suggest that aspects of the response–reinforcer relation may determine the within-session pattern of responding.     

The effects of reinforcer magnitude on timing in rats

The relation between reinforcer magnitude and timing behavior was studied using a peak procedure. Four rats received multiple consecutive sessions with both low and high levels of brain stimulation reward (BSR). Rats paused longer and had later start times during sessions when their responses were reinforced with low-magnitude BSR. When estimated by a symmetric Gaussian function, peak times also were earlier; when estimated by a better-fitting asymmetric Gaussian function or by analyzing individual trials, however, these peak-time changes were determined to reflect a mixture of large effects of BSR on start times and no effect on stop times. These results pose a significant dilemma for three major theories of timing (SET, MTS, and BeT), which all predict no effects for chronic manipulations of reinforcer magnitude. We conclude that increased reinforcer magnitude influences timing in two ways: through larger immediate after-effects that delay responding and through anticipatory effects that elicit earlier responding.     

Effects of long-term shock and associated stimuli on aggressive and manual responses

Squirrel monkeys were exposed to response-independent, fixed-frequency shock that produced biting attack upon a pneumatic hose. Attacks decreased within and across sessions at low intensities and high frequencies of shock, but increased within and across sessions at higher intensities and lower shock frequencies. Stimuli paired with shock, when presented alone, came to produce biting, and stimuli correlated with shock parameters that produced increases in responding within sessions produced similar increases when presented alone. Further experiments showed that continuing exposure to shock also produced lever pressing or chain pulling, with longer shock exposure again producing higher response rates. Whereas biting generally decreased throughout the intershock interval, manual responding generally increased as shock time approached, but immediately before shock was often suppressed. Following shock, biting attack predominated over manual behavior. The results suggest a possible explanation for the extreme resistance of avoidance behavior to extinction, and may also partially explain the persistence of responding during schedules of response-produced shock. Relationships of the present findings to naturalistic observations of relations between fleeing, freezing, and fighting performances are discussed.     

"Turning back the clock" on serial-stimulus sign tracking.

Two experiments examined the effects of a negative (setback) response contingency on key pecking engendered by a changing light-intensity stimulus clock (ramp stimulus) signaling fixed-time 30-s food deliveries. The response contingency specified that responses would immediately decrease the light-intensity value, and, because food was delivered only after the highest intensity value was presented, would delay food delivery by 1 s for each response. The first experiment examined the acquisition and maintenance of responding for a group trained with the contingency in effect and for a group trained on a response-independent schedule with the ramp stimulus prior to introduction of the contingency. The first group acquired low rates of key pecking, and, after considerable exposure to the contingency, those rates were reduced to low levels. The rates of responding for the second group were reduced very rapidly (within four to five trials) after introduction of the setback contingency. For both groups, rates of responding increased for all but 1 bird when the contingency was removed. A second experiment compared the separate effects of each part of the response contingency. One group was exposed only to the stimulus setback (stimulus only), and a second group was exposed only to the delay of the reinforcer (delay only). The stimulus-only group's rates of responding were immediately reduced to moderate levels, but for most of the birds, these rates recovered quickly when the contingency was removed. The delay-only groups's rates decreased after several trials, to very low levels, and recovery of responding took several sessions once the contingency was removed. The results suggest that (a) sign-tracking behavior elicited by an added clock stimulus may be reduced rapidly and persistently when a setback contingency is imposed, and (b) the success of the contingency is due both to response-dependent stimulus change and response-dependent alterations in the frequency of food delivery. The operation of the contingency is compared with the effects of secondary reinforcement and punishment procedures.     

Three factors that promote positive induction when rats respond for 1% sucrose

Studies have demonstrated that rats will increase their operant rate of response for a low-valued reinforcer if a high-valued reinforcer will be available later in the session. Research on this positive induction effect suggests that at least three factors account for its appearance: premature responding for the yet unavailable high-valued reinforcer, an increase in the reinforcing value of the low-valued reinforcer, and responding controlled (e.g., elicited) by the response manipulandum. The present experiment tested whether the size of induction could be systematically altered by varying these factors. Twenty-four rats responded in sessions in which 1% sucrose or a food pellet served as the reinforcer in the first or second half of the session. In some sessions, the same operant response was required in both halves of the session. In others, different responses were required. Half of the rats received the different reinforcers in one food trough while the other half received reinforcers in the different halves of the session in different food troughs. Results demonstrated that a large positive induction effect was observed when all of the above factors were present to contribute to the effect (i.e., high-valued reinforcer upcoming, earned by making the same response, delivered to the same food trough). A small, but significant, induction effect remained when all three were absent (i.e., high-valued reinforcer delivered first, earned by making a different response, delivered to a different food trough). The results support the idea that these three factors are the main contributors to the appearance of this positive induction effect. However, at least one additional factor must also contribute.     

An operant analysis of human altruistic responding

Human altruistic responding (called give responding), which delivered a reinforcer to someone other than the responder, was compared to responding where the responder was the recipient of the reinforcer (called earn responding). The same type of response (button pressing), the same reinforcer (a point representing a penny), and the same reinforcer contingency (a 40-response fixed-ratio schedule) were used for both give and earn responding. Since points representing pennies were used to reinforce give and earn responding, responding for points not worth money was also assessed. Give, earn, and point responding were arranged as concurrent incompatible operants. Lowest rates were obtained for point responding. Compared to earn responding, give responding occurred at lower rates, was more susceptible to cessation when point responding was possible, extinguished more rapidly in the absence of money, and produced less responding during reconditioning compared to conditioning when reconditioning followed a period of nonreinforcement. Give responding was less when it reduced the giver's opportunity to earn. Finally, histories of getting reinforcement from others were shown to determine give responding.