Instructional modeling and the development of visual- and verbal-mediation skills by TMR children.

A study-recall paired-associate (PA) learning task administered to 40 TMR children under one of four instructional-modeling conditions: imagery, verbal mediation, imagery and verbal mediation, and a control condition. On one-half of the PA-learning study trials, the children were provided modeled mediating responses (connective pictures and/or sentences) and on the other trials no model was provided. The children's use of mediating responses on study trials was evaluated as was their recall performance. Each instructional-modeling condition resulted in more effective mediator use and better recall than the control condition. Verbal-mediation training was more effective than instruction in the use of visual imagery. Generation of mediators was most apparent when youngsters were first provided with models and then required to generated their own mediating responses. Gains in mediator use and recall were retained over a period of several days.