The treatment consisted of a combination of edible reinforcers (e.g., candy and popcorn) and verbal praise for correct imitation of the target response or for Spontaneous use (defined initially as response within 2 seconds). However only spontaneous responses were recorded as correct; verbal responses of any other kind were not scored. The time delay was gradually increased by 2-second intervals (up to 10 seconds) as the child responded correctly according to set criteria (e.g. four out of five correct consecutive responses) when a response was mastered training was instituted for the next target behavior in accord with the multiple-baseline across behaviors design. Following training additional sessions were conducted to determine whether the children responded appropriately to five objects (toys) not used in training. Follow-up sessions also were conducted at 1- to 6-month intervals. For each child spontaneous use of target responses increased rapidly (from near zero rates) when the treatment was applied. Each child also responded appropriately when the novel objects were used at the follow-up sessions.