Teaching speech sounds to young children using randomly ordered, variably complex task sequences.

Three children, ages 4:8 to 5:11, were taught a speech sound using a randomly ordered variably complex task sequence (concurrent treatment), in which practice occurred in randomized easy-to-hard tasks. In an AB research design replicated across participants, the children quickly advanced to 60% accurate productions of the target speech sound and displayed generalization to untaught tasks.