Increased Accuracy on One Conditional Discrimination When it Is Linked to a Second Conditional Discrimination

Abstract

The present research investigated whether accuracy on one conditional discrimination could be increased by linking it to a second conditional discrimination. Three pigeons were trained in a discrete-trial procedure. In the first phase of training, a trial had one component, which was an identity delayed matching-to-sample (DMTS) task. In the second phase, a trial had two components. The first was a DMTS task, as before. The second component was a probe. Responding to one probe comparison was reinforced if one sample--comparison had been correct in the prior DMTS component, and responding to the other probe comparison was reinforced if the other sample--comparison had been correct. All three pigeons acquired the conditional discrimination on the probe. In addition, in certain conditions accuracy on the baseline DMTS component was higher in the second phase, with the probe, than in the first phase, without the probe. The present procedure appears promising as a means for assisting individuals with developmental or intellectual disabilities to better come under the discriminative control of the outcomes of their own prior behavior.