Three different types of psychophysical context effects have been studied in comparative experiments with animals and humans. The main context variables investigated were: (1) range of the test series; (2) asymmetry of training to test stimuli (anchor effects); and (3) frequency distribution of the test stimuli. A two-stimulus, two response training procedure, followed by various generalization tests, was used. All subjects (19 chickens and 128 humans) were trained and tested with cubes of different sizes. The psychometric functions support the general assumption that perception in birds undergoes psychophysical context effects similar to that observed in humans. However, while all three variables affected the judgments of human subjects, the choices of chickens and human infants were not strongly affected by the frequency distribution of the test stimuli. These data suggest that two factors are responsible for the three contextual effects investigated: a basic perceptual factor invariant across species and age groups and a cognitive component.

A quantitative frame-of-reference ( FR ) model that has been successfully tested in humans was examined in generalization experiments with chickens. In Experiment 1, three groups of two chickens each were trained to discriminate between cubes different in volume and tested with a series of cubes with volumes either below, surrounding, or above the training stimuli. The obtained psychometric functions support the assumption that asymmetrical testing after two stimulus-two response training leads to the changes predicted by the FR model. In Experiment 2 shifts in the context defining the test series were administered by gradually enlarging the distance between training and test stimuli. While context effects were found in both experiments these were more pronounced in Experiment 2. The results support the general hypothesis that perception in animals undergoes context effects similar to those obtained in human subjects.