Abstract
We compared performance on conditional and trial-unique delayed identity matching-to-sample procedures. In Experiment 1, participants with moderate to severe mental retardation were exposed to both procedures under a single, brief delay value. Three of 5 subjects showed higher accuracy in the trial-unique sessions. In Experiment 2, participants with mild mental retardation were exposed to delay values of 0, 2, 4, 8, and 16 seconds, randomized within each session. For 3 of 4 subjects, accuracy was highest with trial-unique, and lowest on the conditional matching-to-sample, at longer delays. Across the two studies, 6 of 9 subjects showed lower delayed matching accuracy when fewer rather than more stimuli were included in a session.