Effects of feature size and pretraining on the feature - positive effect in pigeons

Abstract

Pigeons were trained on a successive discrimination using pairs of asymmetrical stimulus displays. One display in each pair contained only a feature common to both displays (a white field) while the other contained a distinctive feature (a black line) as well as the common feature. For some pigeons, only responses to the display containing the distinctive feature were reinforced (feature-positive); for the remaining birds, only responses to the display without the distinctive feature were reinforced (feature-negative). The two other variables in the 2 x 2 x 2 design were type of pretraining (one or both displays) and size of the distinctive feature (7% or .5% of display area). The main result was a significant interaction between type of discrimination training and distinctive feature size: in the small feature condition, feature-positive birds performed the successive discrimination more rapidly than feature-negative subjects, but in the large feature condition, the opposite effect was obtained. Peck location recordings made during training suggested that, in the small feature condition, successive discrimination learning was related to the extent of pecking at the distinctive feature. In addition, feature-positive training tended to make pecking at the feature more likely, while feature-negative training had the opposite tendency; although the results were not clear-cut, this was taken to account for the feature-positive effect. There was some evidence that the peck location of subjects in the large feature condition was affected in the same way as that of subjects in the small feature condition. Possible explanations of the failure to obtain a corresponding feature-positive effect in the large feature condition are discussed.