Association for Behavior Analysis International

Coaches make a variety of complex decisions in American-rules college footballespecially related to point-after-touchdown (PAT) conversions and, historically, kicker-style selection. However, little research has characterized the pattern of these choices and whether they are sensitive to environmental manipulations, such as an increase in effort required to score. In the present study, the generalized matching law (GML)a model that predicts a linear relation between choices for two alternatives and the amount of reinforcement garnered from themwas applied to PAT conversions (1 point vs. 2 point) and kicker (soccer-style vs. conventional-style) selection using archived data from the National Collegiate Athletic Association. Overall, both PAT-conversion and kicker selection exhibited matching. Further, narrowing the goal-post width was associated with decreased preference for 1-point PAT attempts and enhanced sensitivity to increases in points scored from 1-point PAT attempts. This investigation provides support for the ecological validity of the GML.

Both human and nonhuman species are found to overexploit resources. This problem has been studied with either public goods games or diminishing returns paradigms. The first studies decision-making under different social environments, and the second studies effects of resource depletion. We combined these two paradigms within an animal model by using a diminishing returns procedure under competitive and noncompetitive social environments. Five pairs of female Sprague Dawley rats were run in adjacent operant chambers separated by a transparent barrier. Each chamber had two levers associated with either a fixed ratio schedule (FR) delivering one food-pellet reinforcer, or a geometric progressive ratio schedule (PR) delivering four reinforcers. Switching from PR to FR reset the PR to its lowest value. Subjects responded independently of each other in a baseline condition, and interactively in a social condition. Subjects generally switched at lower PR values in the social condition, but maintained stable reinforcement rates across both conditions. Of note was the emergence of free-riding behavior in the social condition, where one rat increased the PR value and relied on the partner to renew the resource, resulting in higher reinforcement rates for the free-rider. This behavior, representing resource exploitation, is common in the human situation.

This study investigated the role of contingency between interlocking behavioral contingencies (IBCs) and cultural events (CEs). Three experiments were performed. The task consisted in using the mouse to click on an image presented on the computer screen. In the experiments 1 and 2, individual consequences were programmed for the clicking, cultural consequences were programed for an specific pattern of clicking. In experiment 3, only cultural consequences were programmed. Experiment 1 compared conditions in which CE was presented in a variable interval schedule (VI) against conditions in which it was presented in variable time schedules (VT). Experiment 2 compared VI conditions against VT and extinction (EXT) conditions. Experiment 3 replicated Experiment 2, without individual consequences programmed. The results of these experiments indicate variations of IBCs as a function of the schedules of CE presentations. Extinction effects were observed in EXT conditions and also in VT conditions. In addition, the results of Experiment 3 indicate that a programmed individual contingency is not necessary for the establishment of a metacontingency.

In social foraging, groups of subjects are simultaneously exposed to searching, obtaining and consumption food setting. When several pieces of food are available in a patch, some members of group arrive to gain access and aggressive responses can be observed (Kaspersson, H?jesj? & Pedersen, 2010). This experiment evaluated the relationship between aggressive and consumption responses in groups of pigeons exposed to limited resource whose location was varied. Two groups of pigeons (n=5) were exposed during five sessions to a platform with 12 sealed deposits, only four deposits contained seeds. The piercing seal response delivered seeds and subjects could intake it. For G1, the deposits with food (useful) were located contiguous; for G2 the useful deposits were distant. Each session was composed by two trials; the location of useful deposits was varied in each trial. The sessions were video-recorded, a posteriori were recorded frequency of visits to useful deposits and frequency of aggressive responses (smack, peck about) to each subject. Both groups show that subjects recording higher consumption emit medium aggressive responses; subjects with medium consumption emitted aggressive response al higher level; subjects with lower consumption emitted aggressive responses at low level. Thus, the aggressive responses are related to consumption pattern.

Three pigeons were studied in a token-based accumulation paradigm. Tokens were presented response-independently about every 15 s, according to a variable time (VT) 15 schedule. A fixed ratio (FR) on an exchange-production key stopped the tokens from accumulating and started an exchange phase (signaled by flashing tokens). When pecked during the exchange period, each token produced 3-s access to food. After exchange of tokens, a variable-duration intertrial interval (ITI) occurred. This variable ITI ensured that each trial onset occurred 9 min apart, holding rate of trials constant at 10 per 90-min session. The main independent variable was fixed-ratio (FR) size on the exchange-production key, which varied from 25 to 75 across blocks of sessions. As shown in Figure 1, mean tokens accumulated per trial increased as a direct function of exchange-production FR for all 3 pigeons. These findings are consistent with previous research on the role of FR requirement in a related accumulation/self-control task, and provide a solid baseline against which to assess the role of concurrent behavior in a subsequent phase of the experiment.

During behavioral interventions, practitioners commonly use a variety of reinforces to maintain target behaviors rather than using a single constant reinforcer. However, in basic and applied research, comparing the effects of varied and constant reinforcers has produced mixed results. A probable explanation for such differences is that varied reinforcement is more effective than constant reinforcement only when the delivery of one type of reinforcer increases the reinforcing properties of another; a finding known as complementarity in behavioral economics. In contrast, when one reinforcer reduces the reinforcing properties of another, or functions as substitute, varied reinforcement is no different from constant reinforcement. This hypothesis was tested by comparing the effects of varied and constant reinforcement using complementary and substitute reinforcers on response rate and resistance to change in four adult participants diagnosed with development delays. After preference assessment, the favorite reinforcer, the second reinforcer, or the two were presented using a three-component multiple schedule. In successive conditions the reinforcers were complementary or substitutes. Varied complementary reinforcement produced higher responding than constant reinforcement in three participants but had no effects on resistance to change. These results partially support the notion that complementarity and substitutability play a role when varied reinforcement is used.

Using a concurrent choice paradigm, three pairs of female Long Evans rats were tested in a within-subjects ABA design, with one subject as the focal rat, and the other as the harnessed rat. In the center chamber of a three-chamber apparatus, focal rats were trained to press one lever to access the left chamber, and another lever to access the right chamber. In the baseline condition, focal rats could choose to respond for their partner or an empty chamber and were run until behavior stabilized. In the experimental condition, focal rats could choose to respond for their partner or an unfamiliar rat until stabilization, which was followed by a return to baseline. In the first baseline condition there was a preference for the partner compared to the empty chamber, but a preference for the unfamiliar rat during the experimental condition. The differential responding for familiar and unfamiliar rats between conditions suggests that there is reinforcing value to novel social interaction.

Satisficing is a concept borrowed from economics. It is based on the understanding that all behaviors are choices that require trade-offs between doing things the organism would rather not (pay money, make efforts, miss alternative opportunities) in exchange for obtaining desired things (reinforcers). Satisficing proposes that organisms behave as to obtain satisfactory amounts of the wanted item, by engaging in acceptable amounts of unwanted actions. The key is to determine how satisficing limits are set. Five rats were trained in a token economy to produce and accumulate LEDs (tokens) that could then be exchanged for a fixed volume of a sucrose solution. The initial (token production) link was a concurrent schedule with two response wheels; one under FR and the other under a mixed schedule of positive reinforcement and negative punishment. Both wheels required the same number of responses per reinforcement. Alternatively, rats could switch between the wheels and make combinations. Some combinations initiated the terminal link with lesser responses, thus minimizing effort. With repeated trials at higher schedules, between 70-90% of observed choices were among these satisficing combinations.

This study experimentally investigated human sharing in a laboratory task that simulated environmental variability and resource scarcity (shortfall risk). The project sought to determine whether a risk-reduction model of sharing developed by evolutionary biologists (derived from a risk-sensitive optimization model known as the energy-budget rule) could predict human cooperative behavior. The model predicted that sharing should occur expect when the correlation between the earnings of the participant and partner was not highly positive. Twelve participants responded to earn points exchangeable for money when point gains were unpredictable. Failures to acquire sufficient points resulted in a loss of accumulated earnings (a shortfall). Participants were given the choice between working alone or working with (fictitious) others and then pooling and sharing accumulated earnings. The correlation of earnings between the participant and partner(s) was manipulated across conditions. Results showed that participants chose the sharing option when it was optimal to do so; thereby conforming to the predictions of the risk-reduction model of sharing. However, participants also shared under the positive correlation condition. Thus, participants continued to share although there were no specific benefits for doing so. These results contribute to the understanding of how environmental context and social stimuli influences cooperation and sharing in situations involving risk.

Shaping is an omnipresent process that occurs during the lifetime of the organism and over the evolution of the species. As such, the process of shaping has been the subject of considerable amount of research, some of which has produced a systematic approach. Percentile schedules of reinforcement, for example, have provided researchers with a mathematical equation based on differential reinforcement and the probability of reinforcement. Most often this formula is used to increase some dimension of a response such a frequency or duration. Only a handful of studies have demonstrated the effectiveness of using the percentile schedule to decrease response rates. The present study examines the utility of manipulating response rates with the use of a percentile schedule of reinforcement (m=5, w=0.50) with college student performing a simple computer task using a multi-element design. Preliminary results suggest that percentile schedules of reinforcement are effective at not only increasing but also decreasing response rates.

The SubSearch Game is a videogame in which the player uses a mouse to move a submarine icon on a computer screen in order to retrieve underwater objects. Barriers separating the objects make the task progressively more difficult. The screen is divided vertically in half, with each half containing its own submarine, objects, and barriers. The player can switch between the half-screens at any point. Occasionally, according to concurrent variable-interval variable-interval (VI VI) schedules, the retrieval of an object results in the delivery of points via an on-screen counter and simultaneously in the delivery of coins, which are the players to keep. Retrieval may also produce the loss of points (and the need to return coins to the dispenser) but only if the object is retrieved on the left half of the screen. No losses are scheduled on the right half. Participants played six sessions in which no coin dispenser was used. In two additional sessions (sessions 7 and 8) the coin dispenser was operational. The sessions consisted of six 6-min components in which the reinforcer ratio varied, as did the screen color. Punishers were delivered in half of the components on the left side of the screen. An analysis determined whether there was a significant difference between sessions with points only and with points and coins. Table 1 summarizes the results for 26 participants. The number of clicks was the dependent variable and the number of reinforcers the independent variable. The F-test table shows that the PNlogCLCRa measure of respondingwas significantly affected by punishment and the use of coins.

At the beginning of the experiment, all participants were requested to complete a brief questionnaire. The questionnaire was designed to indicate whether the participant was risk averse or risk seeking. Based on the questionnaire results, participants were assigned to two groups, and the results from each were compared in terms of performance in the SubSearch Game. An analysis determined whether there was a significant difference between the groups. The Sub Search Game is a videogame in which the player uses a mouse to move a submarine icon on a computer screen in order to retrieve underwater objects. Barriers separating the objects make the task progressively more difficult. The screen is divided vertically in half, with each half containing its own submarine, objects, and barriers. The player can switch between the half-screens at any point. Occasionally, according to concurrent variable-interval variable-interval (VI VI) schedules, the retrieval of an object results in the delivery of points via an on-screen counter. Points accumulated during each 36-m session are exchanged for money at its end. Retrieval may also produce the loss of points but only when the object is on the left half of the screen. No losses are scheduled on the right half. The sessions consisted of six 6-min components in which the reinforcer ratio varied, as did the screen color. Punishers were delivered in half of the components on the left side of the screen. The experiment consisted of six sessions. Only the results from the last three were included in the analysis. Table 1 summarizes the results for 26 participants. The number of clicks was the dependent variable and the number of reinforcers the independent variable. The F-test table shows that the PNlogCLCRa measure of respondingwas significantly affected by punishment and risk aversion.

Avoidance is described as behaviour that prevents the occurrence of an aversive stimulus whereas, escape behaviour is described as a behaviour that results in the termination of an ongoing stimulus. There are two types of avoidance, both of which prevent an aversive stimulus from occurring. There is standard avoidance that involves a warning stimulus that signals an aversive stimulus will occur promptly. This type of avoidance is reinforced by the termination of the warning stimulus. The other form of avoidance is called Sidman or free-operant avoidance. This process does not include a warning stimulus and it is unknown what the exact reinforcing properties are that maintain it. Previous research determined that Betta splendens (Siamese fighting fish) do not engage in Sidman avoidance whereas other species including Carassius auratus (goldfish) do. This research looked at whether Betta splendens could be taught Sidman avoidance using prompting and reinforcement strategies. Experimenters established an increase in the frequency of independent crossovers between sides of the experimental tank to avoid an aversive stimulus, these avoidance responses were briefly maintained for two of the subjects and immediately decreased once reinforcement was removed for the third subject.