Experimental self-punishment and superstitious escape behavior

Rats were trained to escape from shock by pressing a bar. Bar holding was subsequently punished with very brief shocks. This treatment failed to depress bar-holding behavior. In some cases, although the escape shocks were delivered very infrequently, bar holding was maintained and resulted in the delivery of several thousand punishments per session. These and other effects of the punishment treatment were investigated. Finally, some of the possibilities of superstitious escape responding were explored by presenting inescapable shocks to rats that had been trained to escape shock by lever pressing. Although responding during these shocks had no programmed consequences, responding was sustained.     

Chained VI performance of pigeons maintained with an added stimulus

Pecks at a first (left) key were reinforced by a partial movement (advance) of the voltmeter behind the key toward the maximum position. When the meter reached maximum deflection, pecks at a second key were continuously reinforced with food. The schedule of reinforcement by which pecking the first key produced a deflection, as well as the size of the deflection, were varied in a series of experiments. The rate of responding on the first key was more affected by changes in the number of steps required to move the meter to the maximum position (the size of the deflection) than by the value of the variable-interval schedule by which pecking produced a meter deflection. The rate at which the bird pecked was low when the meter was near the null position, and high as the meter approached the maximum deflection.     

Escape from an effortful situation

This experiment investigated the tendency to escape from a situation requiring effortful responding. Five human subjects responded in a situation where the response mechanism required 20-lb force to operate; responses were reinforced according to a variable-interval schedule. A subject escaped from this situation by emitting a vocal response which produced a 60-sec “easy period”. During the easy period the reinforcement contingency was switched to a response mechanism requiring 1 lb to operate. It was found that: (1) Escape responding could be conditioned and maintained by producing the easy period; the easy period did not maintain escape responding when the force requirement in the normal situation was equated with it. (2) The rate of escape responding was a function of the magnitude of the force normally required. (3) When easy periods were scheduled after fixed ratios, pausing from the end of the previous easy period to the first escape response was noted. It was concluded that a situation requiring high-force responding is a negative reinforcer. The pattern of fixed-ratio responding suggests that this reinforcer produces typical schedule control in human subjects.     

Delayed escape from light by the albino rat

Two albino rats were trained to terminate an aversive light for 1 min by pressing a bar. After 19 hr of conditioning they were exposed to successive delays of 1, 2, 5, and 10 sec imposed between occurrence of the escape response and light termination. No stimulus change accompanied the delay interval, and any additional responses made at this time reset the delay timer. For both rats the relative frequency of escape responses with very long latencies increased as the delay interval increased. The modal escape latency, however, remained essentially unchanged for all delay values of greater than 1 sec. “Superstitious” responding was observed during the delay interval.     

Some aspects of self aversive stimulation in the hooded rat

Three hooded rats were trained to bar press for variable-ratio liquid reinforcement after which an electric shock was delivered following the response. Initially, the shock was presented on a FR 100 basis but the frequency was gradually increased until all responses were punished. Finally, a partial extinction procedure was conducted to determine if the shock resulted in increased bar pressing. No durable suppression of responding occurred, although one subject's rate was reduced during continuous shock. The overall trend for the three animals was one of increased responding. Changes in the pattern of responding were also observed suggesting that the suppressive effects of the punishment were largely restricted to the first response following reinforcement. Increased responding as a function of shock reintroduction during extinction was also observed.     

Escape responding in a shuttlebox

Measures of response latency are a primary tool for those who investigate escape responding in rats. Unfortunately, the single term “latency to escape” has been applied to several different measures of latency. The present study was designed to demonstrate that two different measures of the “latency to escape” tap different aspects of escape responding. To that end, rats were given escape training using a Sidman Avoidance Schedule with brief inescapable electric shock as the aversive stimulus. The latency to escape the shock-shock interval, as measured from the onset of the last shock in a shock period, did not change across trials. However, the latency to escape the shock period, as measured from the onset of the first shock in a shock period, decreased across trials. In addition, the presentation of a feedback stimulus contingent upon escape responding did not affect the latency to escape either the shock period or a shock-shock interval. The results show that these two latency measures, typically not recognized as unique, measure different characteristics of the strength of escape responding in a shuttlebox. Alternative accounts of this pattern of data were considered.     

Assessment and treatment of problem behavior maintained by escape from attention and access to tangible items.

The results obtained from two consecutive functional analyses conducted with a 6‐year‐old child with autism are described. In the initial functional analysis, the highest rates of problem behavior occurred in the play condition. In that condition, the delivery of attention appeared to occasion problem behaviors. A second functional analysis was conducted wherein an escape from attention condition and a tangible condition were added. In the second functional analysis, higher rates of responding were observed in the escape from attention and tangible conditions. The results suggested that problem behavior was maintained by negative reinforcement in the form of escape from attention and positive reinforcement in the form of gaining access to preferred tangible items. Problem behavior was treated using functional communication training combined with noncontingent reinforcement.     

Variable-interval escape from stimuli accompanied by shocks

Individual performances of three rats were examined under a procedure in which steady rates of bar pressing were maintained by conditioned aversive stimulation. Originally neutral visual and auditory stimuli were accompanied by widely and irregularly spaced pulses of shock; they were terminated on a variable-interval schedule by pressing a bar. The contingencies between behavior and shock were also duplicated in a control procedure in which no visual or auditory stimuli were provided. Pressing observed under the control procedure was attributed to differences in the aversiveness of pressing and nonpressing behavior engendered by differences in the incidence of shock following the two classes of behavior. Increased rates with visual and auditory stimuli were attributed to termination of conditioned aversive stimulation. Control rates declined more rapidly than did experimental rates as the mean interval between successive shocks was lengthened; both rates tended to decline when less than 60 sec was allowed as time out from shocks following the successful response. In the control procedure, discrimination between the continuation and discontinuation of the shock series, as measured by relative rates, depended on the relative length of the interval between shocks and the time-out period. Regular warm-up accelerations in rate were noted following an initial delay in responding at the beginning of each session. The length of time required for the warm-up depended on the length of the mean interval between shocks, indicating that exposure to a certain amount of shock was required to establish a supporting state for the observed performance.     

Characteristics and response-displacement effects of shock-generated responding during negative reinforcement procedures: pre-shock responding and post-shock aggressive responding

Bar-pressing (Experiment I) or key-pressing (Experiments II and III) responses of monkeys were reinforced according to a fixed-interval schedule of negative reinforcement: the first response after a fixed interval of time terminated regularly spaced shocks for a fixed time designated as the reinforcement period. During extinction, shocks continued during the reinforcement period. That there were two types of responding generated by shock alone was indicated by (1) the level of responding maintained during extinction relative to conditions without shock, (2) the stability of two between-shock response patterns across reinforcement and extinction conditions, and (3) the development of these two between-shock patterns without a history of reinforcement. Subjects developed either a pre-shock or a post-shock response pattern when only the bar was available. However, when both a bite tube, an operandum requiring an aggressive topography, and a recessed key, an operandum that did not require an aggressive topography, were provided, the post-shock pattern was observed in tube biting and the pre-shock pattern was observed in key pressing. Removal of the bite tube produced post-shock key responding similar to that observed when only the bar was available. The displacement of post-shock, aggression-motivated responding confirmed the confounding effect of shock-generated responding in negative reinforcement procedures, and suggests that the use of concurrent response alternatives would reduce such confounding.     

The relative influence of place and direction in the Morris water task

Previous work from our laboratory has demonstrated that rats display a preference for directional responding over true place navigation in the Morris water task. The present study evaluated the range of situations in which this preference is observed and attempted to identify methods that favor navigation to the precise location of the escape platform in the room. A preference for directional responding over place navigation was observed in a wide range of procedures that included providing extensive training (Experiment 1), providing only platform placement experience in the absence of active swim training (Experiment 2), training navigation to multiple platform locations in a moving platform variant of the task (Experiment 3), and explicitly training navigation to a precise location in the room, versus navigation in a particular direction, regardless of the pool's position in the room (Experiments 4-5). A modest preference for navigation to the precise spatial location of the platform was observed when the pool wall was virtually eliminated as a source of control by filling it to the top with water (Experiment 6).     

The effects of pre-exposure on escape from a Morris pool

In two experiments rats were pre-exposed to the landmarks surrounding a Morris pool while they swam to a platform with a beacon attached to it. They were then required to escape from the pool by finding the platform, without the beacon, in a new position. When the platform remained in the same place for each pre-exposure session, but was moved from session to session, then subsequent escape from the pool was more rapid than when the landmarks were not visible during pre-exposure (Experiment 1). But when the platform was moved from trial to trial during pre-exposure, then subsequent escape from the pool was disrupted (Experiments 1 and 2). It is proposed that pre-exposure to the landmark alters the attention that is paid to them, which then influences how readily the landmarks can be used to identify the new position of the platform.     

Bar holding during escape conditioning

To analyze bar-holding behavior during escape conditioning three rats were trained to escape from shock by pressing a bar, and three were trained to escape by releasing a bar. Bar holding behavior was stronger and more stable under the release condition. These effects were related to the relative shock duration of onbar versus offbar shocks.     

The time course of extinction-induced aggressive behavior in humans: Evidence for a stage model of extinction

Three experiments examined the effects of extinction on aggressive responding in male college students. In Experiment 1 subjects initially performed on a task where shuttle responding was either continuously or partially reinforced with tokens while either a nonaggressive button-pressing response or an aggressive pad-striking response was concurrently reinforced by escape from a moderately aversive tone. During shuttle acquisition there was clear preference for the escape response of button pressing, but when shuttle responding was extinguished subjects began to respond aggressively by striking the pad to escape. The time course of aggressive escape responses during concurrent shuttle extinction was an inverted U; aggressive responding rose to a peak and then declined. Aggressive responding began earlier in extinction following continuous- as contrasted with partial-reinforcement shuttle training. Experiment 2 showed that similar extinction-induced aggression was precipitated by both moderate and extended continuous-reinforcement shuttle training, with earlier onset after extended reinforcement. Experiment 3 ruled out the possibility that the emergence of the pad-striking response during extinction was simply induced response variation. These data were interpreted within the theoretical framework of P.T.P. Wong's recently advanced stage model of extinction (Animal Learning and Behavior, 1978,6, 82–93).     

Fixed-ratio escape and avoidance-escape from naloxone in morphine-dependent monkeys: effects of naloxone dose and morphine pretreatment.

Lever pressing by rhesus monkeys was maintained by morphine injections during four equally spaced sessions each day. During other periods, lever pressing was maintained by timeout from a continuous naloxone infusion (escape), or by timeout from a stimulus that preceded naloxone injections, or termination of the injections (avoidance-escape). As naloxone dose increased in the escape procedure, response rate increased to a maximum and then decreased. In the avoidance-escape procedure, response rate generally increased as naloxone dose increased, but the changes in rate were small compared to the excape procedure. Substitution of saline for naloxone in the escape procedure led to a very low response rates within three sessions. In the avoidance-escape procedure, rate decrements produced by saline substitution appeared to be related to the behavioral history of the monkey. Previous escape experience led to more rapid decreases in responding when saline was introduced, whereas responding was maintained for 15 sessions in a monkey without prior escape conditioning. Morphine pretreatment produced comparable, dose-dependent decreases in response rates in both procedures. The rate-decreasing effects of morphine were exacerbated when no naloxone was delivered in the escape procedure.     

The effects of escape conditioning and shock intensity on responding during inescapable shock

Eight albino rats, conditioned to press a lever to escape shock, continued to lever press during short inescapable shocks presented subsequently. The rate of this behavior was found to be higher for higher shock intensities regardless of the order in which shock values were presented. Relative to the immediately preceding escape rate, responding during inescapable shock was higher following conditioning at higher fixed-ratio escape requirements. Four subjects not conditioned to escape shock pressed the lever very infrequently during inescapable shock and showed little change with changes in shock intensity. The escape conditioning effects suggest that responding during inescapable shock is superstitious escape behavior. The effects of shock intensity on this behavior appear to be similar to reported effects of shock intensity on escape behavior.     

The effects of dimensional naming upon children’s performance in a modified optional shift problem

In a modified optional shift paradigm, shift and test tasks were administered concurrently to 120 second-grade children. Ss required, during shift learning, to verbalize the values of the previously relevant dimension showed an increase in reversal responding in the test task, whereas those required to verbalize the values of the previously irrelevant dimension showed an increase in nonreversal responding. The results are in good agreement with predictions made from an extension of Hull-Spence discrimination learning theory.     

Effects of a concurrent task on fixed-interval responding in humans

Subjects pressed a telegraph key to illuminate a meter dial on which pointer deflections appeared at fixed intervals. Upon detecting a deflection they were required to press another key to reset the pointer to zero. This detecting and resetting operation reinforced the behavior of pressing the light-flashing key (i.e., the observing responses). The usual pattern of responding on the light-flashing key was a long pause following the reinforcement and an abrupt transition to a steady response rate toward the end of the interval. When the subjects were required to perform a concurrent subtraction task, the pattern of responding changed in varying degrees, ranging from complete loss of typical fixed-interval behavior to a slight shortening of the post-reinforcement pause. These effects were attributed to the disruption of the self-produced verbal chains (counting or reciting) that ordinarily govern human behavior on this schedule.     

The effects of control and lack of control on active and passive avoidance in rhesus monkeys

Most studies which have examined the effects of lack of control have utilized test tasks in which active responding is required, and generally they have found impaired learning. Those few studies which have required passive responding in the test task generally have found facilitation of learning. The present two experiments examined the effects of lack of control in both active and passive avoidance tasks in a primate species (Macaco mulatta) not previously used in this research area. In Experiment 1, although the group without control (IE) tended to be inferior at active and superior at passive avoidance in comparison to the group with control (E), there were no significant differences. In Experiment 2, utilizing a difficult discrimination task in which subjects were required to learn when and when not to respond actively to avoid aversive stimulation, greater group differences were found. Two monkeys from Group IE failed to escape in active avoidance acquisition and, as a whole. Group IE was somewhat slower to respond than Group E. At passive avoidance, however. Group IE was superior to Group E and, as a consequence, more efficiently solved the discrimination problem. Implications of the present results for interpretation of the effects of lack of control as deficits are discussed.     

The sunk cost effect with pigeons: some determinants of decisions about persistence

The sunk cost effect occurs when an individual persists following an initial investment, even when persisting is costly in the long run. The current study used a laboratory model of the sunk cost effect. Two response alternatives were available: Pigeons could persist by responding on a schedule key with mixed ratio requirements, or escape by responding on a second key. In Experiment 1, mean response requirements for persistence and escape were varied across conditions. Pigeons persisted (committing the sunk cost error) when persisting increased the mean response requirement only slightly but not when persisting was sufficiently nonoptimal. Experiment 2 explored more systematically combinations of ratios and probabilities assigned to the schedule key. Persistence varied with the ratio of the mean global response requirements for persistence and escape. In Experiment 3, transitions between ratios were signaled. This reduced nonoptimal persistence, and produced some instances of a reverse sunk cost error--escaping when persistence was optimal. In Experiment 4, it was optimal to escape after the second-smallest ratio ever presented. Pigeons escaped at approximately the optimal juncture, especially in conditions with added signals. Overall, this series of experiments suggests that the sunk cost error may arise in part because persistence is the default behavioral strategy in situations where the contingencies for escape and persistence are insufficiently disparate and/or it is relatively difficult to discriminate when to escape. The study also demonstrates the utility of animal models of complex decision making situations.