Reduced versus nonreduced models in language training of MR children

Two groups of 5 MR children were given language training on a highly structured training program. During training, one group received only reduced (telegraphic) imitation models, while the other group received only nonreduced imitation models. Results indicated that language training with reduced models increased the performance of Ss during the lessons and on posttesting. Performance differences were also found to generalize from the use of pictures as stimulus materials to the use of real objects. Neither form of language training was found to influence Ss comprehension of the concepts taught.