Association for Behavior Analysis International

Assessing the preferences of animals in human care is important for many reasons. Practically speaking, an understanding of animals' preferences can help caregivers predict response patterns and enhance training and enrichment strategies. The presentations in this symposium describe efforts to apply behavior-analytic methodology to the assessment of animal preferences. Subjects include domestic and exotic animals, and formats include MSWO and PSPA preference assessments, as well as an assessment utilizing a disequilibrium perspective of reinforcer valuation. Dr. Meagan Gregory will bring her expertise in preference assessment methodology and application to bear in an analysis of the work presented.

This event is appropriate for behavior analysts who want to learn more about applying preference assessment methodology in applied animal behavior contexts.

Husbandry and medical training are critical components of animal care in zoos. Training allows animals to voluntarily participate in their own care in many ways, including day-to-day husbandry routines such as shifting and visual examinations as well as periodic medical procedures such as injections and ultrasound examinations. Maintaining these behaviors throughout an animal's life requires careful selection of reinforcers, which might be aided by the use of a preference assessment protocol. We describe the development of a multiple stimulus without replacement (MSWO) preference assessment protocol for regular use with African lions, an attempt to validate the protocol by comparing preference test choices with training trial outcomes, and next steps in applying this evidence-based practice in a new setting.

The paired-stimulus preference assessment (PSPA) has been widely accepted as the gold standard for identifying preferred items. These are then used in teaching to promote efficient learning. We examine the application of the PSPA to a group of black vultures, and discuss similarities and differences in responding across members of the same species. Applied animal training should incorporate more of these well-known behavior-analytic procedures to improve the training and quality of life for animals in captivity.

Animals in Shelters need training and/or enrichment, whether they are a social species, such as a dog or cat, or whether they are an animal that lives with humans, such as a rat or a parrot. Training can make a dog (and sometimes a cat) more adoptable, and training can be seen as an actual form of enrichment for an animal that does not live in a home. An enriched environment is one in which an animal has variety, choice and control over its daily activities, and environmental enrichment can be described as an improvement in the biological functioning of captive animals resulting from modifications to their environment. Preference assessments fulfill a need in both helping trainers and handlers to choose possible reinforcers and to discover preferences of possible enrichment food or materials. In this experiment, as array of known preference assessments were applied to a wide array of animals in an open-admission shelter; data showed that nearly all animals made choices when given the opportunity.

The Disequilibrium Model (Timberlake & Farmer-Dougan, 1991) states that the degree to which the ratio of instrumental (I) to contingent (C) responding is disrupted from a baseline bliss point (Oi/Oc) results in predictable reinforcement effects. When the contingency reduces the contingent response below baseline (I/C > Oi/Oc), a deficit is imposed for the contingent response (e.g., treat eating), and the rate of instrumental responding (e.g., initiate to human) increases. Punishment effects occur when the I/C is pushed above baseline, I/C < Oi/Oc, producing a satiation in operant responding. For example, dogs with little experience with human interaction may have different Oi/Oc baselines than experienced dogs when human-delivered treats are used to reinforce responding to human cues. The baseline rate of human approaches for inexperienced dogs may be lower, thus a contingency which requires increased approaches may result in reduced approaches to humans: a punishment effect, I/C > Oi/Oc). This paper presents data demonstrating predicted differences in reinforcer efficacy between dogs who were experienced with or those with little experience interacting with humans. The data support the predictions of the disequilibrium model, and demonstrate the importance of assessing baseline rates of both the contingent and operant response to determine reinforcer efficacy.