The general goals of this research are to determine the cognitive processes that take place when decisions are made. The decision processes will be studied in well-learned tasks under different conditions of difficulty, and in new tasks which the subjects are learning. The subjects in these experiments will be pigeons and humans (both adults and children). The matching-to-sample task will serve as a model environment for the study of these decision processes. A sample stimulus (color) is presented and in a following test the subject must chose the one that matches the sample (matching) or is different from the sample (oddity). In some situations the sample is present when the choice is made (simultaneous), and in others it has been removed (delayed). Preliminary research has identified the decision strategies that pigeons use to solve this task. They consider each choice possibility individually either accepting it or rejecting it as the sample match. If they reject both, they return to the first and repeat the process anew as if it had never been seen before, and so on. This decision strategy is in mathematical terms a Markov process. It is a rare opportunity to be able to study a behavioral Markov process. Such processes are very powerful mathematically and can serve as a benchmark against which to study other decision strategies. The establishment and verification of the strategy is based upon observable responses (movements to see the stimuli), and the outcome is the identification of cognitive (decision) strategies. How subjects learn and apply this decision rule will be studied in a variety of learning and transfer experiments, and the role of the sample and the latitude of their """"""""acceptance"""""""" will be explored in a variety of experiments which they are manipulated. The human experiments provide a human reference and comparison to these very basic decision rules employed by pigeons.
