Acquisition of Same and Different Choice Behavior

The panel below shows mean performance on Same and Different trials collapsed across the four display types. It shows that accurate choice responding for these two trial types emerged at different points during training, with increased correct responding on Same trials generally occurring about 15 to 20 sessions prior to an increase on Different trials. The lower panels show this same data broken down by the four display types.

We believe that the inherently smaller number of Same displays that can be formed in any Same-Different discrimination may be responsible for the initially higher accuracy observed with Same trials. Despite their considerable number, the Same displays had to be repeated more frequently than Different displays in the present experiment. This repetition may have encouraged the birds to initially try and memorize their responses to these specific displays. Depending on how you count them, their quantity was within the range of items that pigeons have been shown to be able to memorize (Vaughan & Greene, 1984). Of course, this memorization strategy would fail, or at least be very difficult and time consuming, when applied to the other half of the present discrimination involving the very large number of Different displays. This difficulty may have forced the birds to give up on this first approach and switch to a more conceptual-based strategy at a later point in training.