The relation between response and attentional shifts in pigeon compound matching-to-sample performance

Pigeons’ pecks to the two elements of spatially separated compound samples were observed during matching-to-sample performance. An attentional biasing procedure was used in which the birds were tested on a subset of the information contained in the sample (either one of the two dimensions or one of the two sample key locations) for a large number of sessions. This procedure resulted in a greater proportion of sample pecks to the positively biased (tested during biasing) sample attribute. Increases in accuracy on the positively biased attribute as well as decreases in accuracy on the negatively biased (not tested during biasing) attribute were also found. However, not all of the shifts in matching accuracy could be explained by the shifts in pecking behavior.