seen and inspected the door. Experienced technicians wait until an animal visually fixates on the escape door before encouraging it to leave its home cage. Second, training should proceed in steps with attainable goals established. Once begun, it should continue until an animal has accomplished the goal for that session.

In avoidance training, an animal is encouraged to escape from or avoid something that it considers undesirable. The animal should be allowed to "escape" the threat of a net, gloved hand, or cage squeeze no matter how easily it might be caught at the time. The threat of noxious stimuli is more effective than the actual use of such stimuli. Escape from threat of restraint by a gloved hand or net is more effective than escape from restraint. The animal should be able to control the perceived threat of noxious stimuli by using an effective means of escape. Avoidance-training procedures should discontinue the aversive stimulus immediately. For example, if a squeeze-back cage is used to immobilize an animal's arm or leg for examination, the squeeze back should be released as soon as the limb is obtained; maintaining the squeeze while one examines the limb is counterproductive, in that no connection is established between presenting the limb and having restraint removed. Similarly, if resistance is rewarded by the technician's giving up, it will become even harder to overcome later the "reward" of the first experience. For example, an animal can learn that racing around a cage aimlessly can lead to withdrawal of the technician, net, or hand. Training should continue until the session's goal is achieved.

That principle also applies to positive reinforcement. If a favorite food is to be offered when an animal enters a new cage, giving the same food to a resisting animal rewards resistance rather than compliance. With patience, many animals can be trained to comply with laboratory requirements through the use of positive reinforcement alone. Reinforcement should be prompt. When an animal has complied with the desired activity, positive reinforcement (for example, food reward) should follow immediately. There should be no delay while notes are written or other activities performed.

Third, the procedure should be routine. When animals are to be handled or restrained, technicians and researchers should wear distinctive clothing, such as a different-colored laboratory coat or clothing other than what is worn for routine feeding and watering, behavioral observation, research, or any other daily activity. Although the animals might still recognize the individuals participating in the activity, the distinctive clothes separate this somewhat invasive situation from all other daily routines. Quiet, deliberate movements result in more effective cooperation than noisy, abrupt activity. Use of a routine can reduce overall stress. For example, capture of animals in a fixed order will allow animals to learn when their turn is coming and produce less stress than a varying order of capture and handling.

Fourth, animals should be allowed some latitude in performance. Each primate is an individual and can respond to a given situation idiosyncratically.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement