Time-delay discrimination training: replication with different stimuli and different populations.

Two time-delay conditions for teaching complex visual discriminations to normal preschoolers and children with mild and moderate intellectual handicaps were compared. One condition involved spatially separating the distinctive components from the redundant parts of both stimuli (multiple dynamic distinctive-feature prompts). The other condition involved adding a colored field to the correct stimulus (single static nondistinctive-feature prompt). The effect of the latter condition was assessed with unlearned and learned tasks. The study consisted of four experiments. In one experiment, children were also required to use the prompts for self-monitoring responses given before prompting had occurred. The results indicated that for all populations and stimuli (a) time delay of multiple dynamic distinctive-feature prompts consistently produced learning, and (b) time delay of the single static nondistinctive-feature prompt almost never produced learning and frequently led to a complete loss of discriminative performance on previously learned tasks. The resistance to disruption was a function of the training history (i.e., with or without time delay) and IQ level. Self-monitoring increased the efficacy of time delay when multiple dynamic distinctive-feature prompts were used but not when the single static nondistinctive-feature prompt was used.